During a laparoscopic surgery, an endoscopic camera is used to provide visual feedback of the surgery to the surgeon and is controlled by a skilled assisting surgeon or a nurse. However, in robot-assisted teleoperated systems such as the daVinci surgical system, the same control lies with the operating surgeons. This results in an added task of constantly changing view point of the endoscope which can be disruptive and also increase the cognitive load on the surgeons. The work presented in this thesis aims to provide an approach that results in an intelligent camera control for such systems using machine learning algorithms. A particular task of pick and place was selected to demonstrate this approach. To add a layer of intelligence to the endoscope, the task was classified into subtasks representing the intent of the user. Neural networks with long short term memory cells (LSTMs) were trained to classify the motion of the instruments in the subtasks and a policy was calculated for each subtask using inverse reinforcement learning (IRL). Since current surgical robots do not enable the movement of the camera and instruments simultaneously, an expert data set was unavailable that could be used to train the models. Hence, a user study was conducted in which the participants were asked to complete the task of picking and placing a ring on a peg in a 3-D immersive simulation environment created using CHAI libraries. A virtual reality headset, Oculus Rift, was used during the study to track the head movements of the users to obtain their view points while they performed the task. This was considered to be expert data and was used to train the algorithm to automate the endoscope motion. A 71.3% accuracy was obtained for the classification of the task into 4 subtasks and the inverse reinforcement learning resulted in an automated trajectory of the endoscope which was 94.7% similar to the human trajectories collected demonstrating that the approach provided in thesis can be used to automate endoscopic motion similar to a skilled assisting surgeon.

A series of sol-gel based photopolymerizable silica glass for holographic storage have been synthesized via a facile chemical design in an effort to overcome the cracking problem incurred during the irradiation stage. The holographic composite material includes a low molecular weight polydimethylsiloxane (PDMS) with end-capped hydroxyl groups, an epoxide-containing coupling agent, a photopolymerizable acrylate monomer, and the sol-gel-derived silica matrix. Inclusion of PDMS provides improved compression stress and strain and toughness over the original unmodified samples, without deterioration on the diffraction efficiency (£b). A plateau value of £b ~ 40 % can be found under a beam power of 5.02 mW. Correlation between the molecular weight of acrylate polymer and £b of the holographic composite material is described and the discrepancies analyzed. The premature saturation of polymerization in the holographic system is attributed to the retarded free radical polymerization within the silica matrix. Minor variation of £b with photoinitiator concentration is observed, indicating that only moderate molecular weight is required for a reasonably high £b. The crack-free mechanism is based on the chemical and physical interactions within the material system. We also investigated the pore size and volume effect for optical properties, especially in studies of diffraction efficiency. With mixing rubber additives, the pore characteristics change via measurement of BET technique. The results indicated that moderate pore size made high diffraction efficiency for porous matrix and some characterizations also were presented. On the other hand, the optical application of this recording medium was described via 3D projected fringes profilometry (PFP) measurement. We recorded coded pattern on medium by employing fiber devices as light source. Using endoscope as image-derived tool, we get double-projected fringe images on tiny object. By PFP technique, a performed retrieved image can be achieved.

This dissertation presents the design of a miniature injection-molded objective lens for a fiber-optic confocal reflectance microscope. This is part of an effort to demonstrate the ability to fabricate low cost, high performance biomedical optics for high resolution in vivo imaging. Disposable endoscopic microscope objectives could help in vivo confocal microscopy technology mature to enable large-scale clinical screening and detection of early cancers and pre-cancerous lesions. This five lens plastic objective has been tested as a stand-alone optical system and has been coupled to a confocal microscope for in vivo imaging of cells and tissue. Changing the spacing and rotation of the individual optical elements can compensate for fabrication inaccuracies and improve performance. An optical-bench testing system was constructed to allow interactive alignment during testing. The modulation transfer function (MTF) of the miniature objective lens is determined using the slanted-edge method. A custom MATLAB program, edgeMTF, was written to collect, analyize, and record test data. An estimated Strehl ratio of 0.64 and an MTF value of 0.70, at the fiber-optic bundle Nyquist frequency, have been obtained. The main performance limitations of the miniature objective are mechanical alignment and flow-induced birefringence. Annealing and experimental injection molding runs were conducted in effort to reduce birefringence.

abstract: The objective of this project was to evaluate human factors based cognitive aids on endoscope reprocessing. The project stems from recent failures in reprocessing (cleaning) endoscopes, contributing to the spread of harmful bacterial and viral agents between patients. Three themes were found to represent a majority of problems: 1) lack of visibility (parts and tools were difficult to identify), 2) high memory demands, and 3) insufficient user feedback. In an effort to improve completion rate and eliminate error, cognitive aids were designed utilizing human factors principles that would replace existing manufacturer visual aids. Then, a usability test was conducted, which compared the endoscope reprocessing performance of novices using the standard manufacturer-provided visual aids and the new cognitive aids. Participants successfully completed 87.1% of the reprocessing procedure in the experimental condition with the use of the cognitive aids, compared to 46.3% in the control condition using only existing support materials. Twenty-five of sixty subtasks showed significant improvement in completion rates. When given a cognitive aid designed with human factors principles, participants were able to more successfully complete the reprocessing task. This resulted in an endoscope that was more likely to be safe for patient use. / Dissertation/Thesis / M.S. Applied Psychology 2011

Introdução: Biofilme consiste em um conjunto de microrganismos embutidos em uma matriz extracelular. O biofilme tradicional (TBF) se desenvolve em locais com contínua hidratação, e o biofilme buildup (BBF), em produtos para a saúde expostos a repetidos ciclos de sujidade, limpeza, desinfecção e secagem. Objetivo: avaliar o impacto do uso de detergente e fricção para remover biofilme buildup-5 (BBF-5) e biofilme tradicional-5 (TBF-5) de canais endoscópicos gastrointestinais flexíveis, bem como o impacto do glutaraldeído após a limpeza. Métodos: O BBF-5 foi desenvolvido após a exposição da superfície interna dos canais de endoscópicos gastrointestinais ao Artificial Test Soil (ATS) contendo 108UFC/ml de Pseudomonas aeruginosa (PA) e Enterococcus faecalis (EF), limpeza, desinfecção de alto nível (DAN), enxágue e secagem durante cinco dias. O mesmo processo foi utilizado para desenvolver o TBF-5, porém a fase de DAN foi omitida. Limpeza com detergentes com enzimas de pH neutro e detergentes sem enzimas com pH alcalino, escova com cerdas e dispositivo de limpeza pull thru foram comparados à limpeza com água e sem fricção para remover o BBF-5 e o TBF-5. Testes de proteína e carboidrato, contagem de bactérias viáveis e adenosina trifosfato (ATP) foram realizados. Teste de Kruskal-Wallis e análise de post hoc Dunn-Bonferroni foram realizados. Resultados: Não houve diferença estatisticamente significante (p>0,05) entre a quantidade de EF e PA nos canais com TBF-5 e BBF-5. Não houve diferença da eficácia para a remoção de PA em TBF-5 entre o detergente enzimático e a água, em ambas as situações, a limpeza apenas não foi eficaz (p>0,05) na ausência de fricção (flush). Por outro lado, o detergente alcalino, associado a todos os procedimentos de limpeza (ausência e presença de fricção) e DAN, foi eficaz para remoção de PA. Para a remoção de EF em TBF-5, a eficácia da limpeza foi alcançada em todas situações avaliadas (com e sem fricção) com o uso do detergente enzimático; com a água, na presença de fricção (pull thru e escovas de cerdas); e com detergente alcalino, apenas com o método de limpeza com escovas de cerdas (p<0,05). Para a remoção de bactérias em BBF-5, formados em canais de PTFE, não houve diferença da eficácia para a remoção de PA entre o detergente enzimático e o alcalino, em ambas situações, todos os procedimentos avaliados foram eficazes (p<0,05), enquanto que a água foi ineficaz para remoção de PA apenas na ausência de fricção (p>0,05). Para a remoção de EF em BBF-5, o detergente enzimático e a água foram eficazes em todas situações avaliadas; o detergente alcalino, foi eficaz na presença de fricção com o dispositivo pull thru e escovas de cerdas (p<0,05). Testes de ATP, proteína e carboidrato foram incapazes de detectar biofilme. Conclusão: Água potável de torneira, detergente com enzimas e detergente alcalino sem enzimas foram eficazes para a completa remoção de BBF-5 e TBF-5 na presença de fricção (pull thru ou escova com cerdas) durante a limpeza. Na ausência de fricção, detergente enzimático apresentou maior habilidade para remover E. faecalis de TBF-5 e BBF-5, detergente alcalino para remover P. aeruginosa e a água não foi eficaz para a remoção de ambos microrganismos. Glutaraldeído destruiu bactérias remanescente após a maioria das combinações de limpeza avaliadas. / Background: Biofilm is the accumulation of microorganisms enclosed in an extracellular matrix. Traditional biofilm (TBF) develops under continuous hydration whereas buildup biofilm (BBF) develops on medical devices that are exposed to repeated cycles of soiling, cleaning, disinfection and drying. Objetive: to evaluat the impact of detergent and friction on BBF-5 and and TBF-5 removal from PTFE channels as well as the killing efficacy of glutaraldehyde post cleaning. Methods: BBF-5 was developed by repeated rounds over five days of exposure to bacteria, cleaning, high level disinfection (HLD), rinsing and drying of the inner surface of new polytetrafluorethylene (PTFE) channels. Artificial Test Soil (ATS) containing 108CFU/ml of Pseudomonas aeruginosa (PA), and Enterococcus faecalis (EF) was used for the bacterial exposure. The same process was used to develop TBF-5, however the HLD step was omitted. Cleaning with enzymatic and alkaline detergents, bristle brush and pull thru cleaner were compared to a water-flush only to determine BBF-5 and TBF-5 removal. The residual organic and microbial levels were tested using carbohydrate, and protein assays as well viable count, and adenosine triphosphate (ATP) testing. Kruskal-Wallis test and the post hoc Dunn-Bonferroni analysis were performed. Results: There was not a statistically significant difference (p> 0.05) between the amount of EF and PA in the channels with TBF-5 and BBF-5. There was not a difference in efficacy for the removal of PA in TBF- 5 between the enzymatic detergent and water and in both situations, only cleaning was not effective (p>0.05) in the absence of friction (flush). Furthermore, the alkaline detergent, associated with all cleaning procedures (absence and presence of friction) and DAN, was effective for the removal of PA. For removal of EF in TBF-5, cleaning efficacy was achieved in all procedures evaluated (with and without friction) with the use of enzymatic detergent; with water in the presence of friction (pull thru and bristle brushes); and an alkaline detergent, the only method of cleaning brushes with bristles (p<0.05). For the removal of bacteria in BBF-5 formed on PTFE channels, there was not a difference in efficacy for PA removal of the enzymatic and alkalyne detergent and in both situations, all procedures were evaluated effective (p<0.05 ), while water was ineffective for PA removal only in the absence of friction (p>0.05). For EF removal on BEF-5, the enzymatic detergent and water were evaluated effective in all situations; alkaline detergent, was effective in the presence of friction with the pull thru device and bristle brushes (p<0.05). ATP, protein and carbohydrate testing were unable to detect biofilm. Conclusion: Tap water, enzymatic and alkaline detergent were effective for complete BBF-5 and TBF-5 removal in the presence of friction (pull thru or bristle brush) and a flushing pump during the cleaning. Without friction, enzymatic detergent was more effective for EF removal of TBF-5 and BBF-5, alkalyne detergent was more effective for PA removal and water was not effective for both types of bacterial removal. Glutaraldehyde effectively killed the remaining microorganisms after some cleaning combinations were tested.

A multi-modality imaging system intended for clinical utilization has been developed. It is constructed around an existing fiber-bundle-based fluorescence confocal microendoscope. Additional imaging modalities have been implemented to expand the capabilities of the system and improve the accuracy of disease diagnosis. A multi-spectral mode of operation is one such modality. It acquires fluorescence images of a biological sample across a spectral range of sensitivity and explores the collected image data at any specified wavelength within that spectral range. Cellular structures can be differentiated according to their spectral properties. The relative distribution and concentration of the different cellular structures can potentially provide useful pathologic information about the imaged tissue. A spectral-domain optical coherence tomography (SDOCT) modality is another imaging technique integrated into the system. It provides a cross-sectional imaging perspective that is comparable to microscopic images obtained from histology slides and complementary to the en face view obtained from the confocal imaging modality. The imaging system uses a parallelized architecture (fiber-optic bundle, line of illumination) to increase the data acquisition speed. A one-dimensional scan is needed to capture 2D images in the confocal modality or a 3D data cube (two spatial dimensions and one spectral dimension) in the multi-spectral mode of operation. No scanning is required to capture a 2D OCT image. The fiber-bundle design is particularly critical for the SDOCT modality as it paves the way to novel fast endoscopic OCT imaging that has a high potential for translation into the clinic. The integrated multi-modality imaging system can readily switch between different imaging modalities, which will make it a powerful diagnostic tool in a clinical environment. It can provide valuable information about the morphology, the spectral and biochemical features, and the macroscopic architecture of tissue. It is believed that fast and accurate disease diagnosis can potentially be made based on all these characteristics.

The optical design for a dual modality endoscope based on piezo scanning fiber technology is presented including a novel technique to combine forward-viewing navigation and side viewing OCT. Potential applications include navigating body lumens such as the fallopian tube, biliary ducts and cardiovascular system. A custom cover plate provides a rotationally symmetric double reflection of the OCT beam to deviate and focus the OCT beam out the side of the endoscope for cross-sectional imaging of the tubal lumen. Considerations in the choice of the scanning fiber are explored and a new technique to increase the divergence angle of the scanning fiber to improve system performance is presented. Resolution and the necessary scanning density requirements to achieve Nyquist sampling of the full image are considered. The novel optical design lays the groundwork for a new approach integrating side-viewing OCT into multimodality endoscopes for small lumen imaging.

Wireless video capsule endoscopy (VCE) plays an increasingly important role in assisting clinical diagnoses of gastrointestinal (GI) diseases. It provides a non-invasive way to examine the entire small intestine, where other conventional endoscopic instruments can barely reach. Existing examination systems for the VCE cannot track the location of a endoscopic capsule, which prevents the physician from identifying the exact location of the diseases. During the eight hour examination time, the video capsule continuously keeps taking images at a frame rate up to six frame per sec, so it is possible to extract the motion information from the content of the image sequence. Many attempts have been made to develop computer vision algorithms to detect the motion of the capsule based on the small changes in the consecutive video frames and then trace the location of the capsule. However, validation of those algorithms has become a challenging topic because conducting experiments on the human body is extremely difficult due to individual differences and legal issues. In this thesis, two validation approaches for motion tracking of the VCE are presented in detail respectively. One approach is to build a physical testbed with a plastic pipe and an endoscopy camera; the other is to build a virtual testbed by creating a three-dimensional virtual small intestine model and simulating the motion of the capsule. Based on the virtual testbed, a physiological factor, intestinal contraction, has been studied in terms of its influence on visual based localization algorithm and a geometric model for measuring the amount of contraction is proposed and validated via the virtual testbed. Empirical results have made contributions in support of the performance evaluation of other research on the visual based localization algorithm of VCE.

A técnica cirúrgica endoscópica transluminal por orifícios naturais (NOTES) tem por definição envolver a introdução de um endoscópio no interior da cavidade abdominal através de uma perfuração intencional de uma víscera (p.ex. estômago, reto, vagina) e realizar uma exploração e/ou operação intra-abdominal. A incisão cirúrgica realizada pela técnica de NOTES tem a vantagem de ser utilizada para a retirada da cavidade abdominal de espécimes (NOSE) reseccionadas através de uma laparoscopia padrão ou por NOTES. O presente estudo teve por objetivo verificar a viabilidade da cânula vaginal em realizar o acesso transvaginal para a cavidade abdominal através da técnica de NOTES transvaginal em éguas hígidas posicionadas em estação e avaliar o acesso vaginal como canal de extração de espécimes (NOSE) em éguas submetidas à ovariectomia laparoscópica via flanco. Foram utilizadas 6 éguas hígidas para ambos os procedimentos cirúrgicos. Os animais foram submetidos a jejum alimentar pré-cirúrgico de 24 à 30 horas. Para a sedação utilizou-se cloridrato de detomidina e cloridrato de lidocaína para execução da epidural baixa. A incisão vaginal foi realizada dorsal à cérvix na posição horária de 12 horas utilizando uma cânula vaginal de ponta romba rosqueada e realizada sob visualização indireta através do endoscópio flexível inserido dentro da cânula. A remoção dos ovários via vaginal ocorreu em quatro das seis éguas utilizadas no experimento. Das quatro éguas, três tiveram a necessidade do aumento manual da colpotomia para apreensão manual dos ovários intracavitária e sua remoção. Em dois animais os ovários tiveram de ser seccionados e removidos através do flanco pelo aumento da incisão do canal de trabalho laparoscópico. A execução cirúrgica do acesso à cavidade abdominal via transvaginal e a técnica de extração de espécimes por NOSE submetidas à ovariectomia laparoscópica via flanco é possível de ser realizada em éguas hígidas. / The technique of Natural Orifice Transluminal Endoscopic Surgery (NOTES), by definition, concerns the introduction of an endoscope in the abdominal cavity through an intentional perforation of a viscera (for ex., stomach, recto, vagina) and performs an intra-abdominal exploration and/or operation. The surgical incision performed through the NOTES technique has the benefit of being used to remove resected specimens (NOSE) from the abdominal cavity through a standard laparoscopy or through NOTES. The present study had the purpose to check the practicality of the vaginal cannula in performing transvaginal access to the abdominal cavity through the NOTES technique in standing healthy mares and assess the vaginal access as an canal extraction of specimens (NOSE) in mares subjected to laparoscopic ovariectomy. Six healthy mares were used for both surgical procedures. The food was withheld between 24 to 30 hours before surgery. They were sedated with detomidine hydrochloride and to perform the low epidural, lidocaine hydrochloride was used. The vaginal incision was performed dorsally to the cervix at 12 o’clock position using threaded round tip vaginal cannula and performed under indirect viewing through a flexible endoscope inserted in the cannula. The ovaries were removed through the vagina in four of the six mares used in the experiment. From the four mares, three were required to manually increase the colpotomy for manual intracavitary seizure of the ovaries and their removal. In two of the animals, the ovaries had to be cut and removed through the flank, increasing the incision of the laparoscopic working channel. The procedure to create the access to the abdominal cavity through the vagina to perform the NOTES technique and the specimen extraction technique through NOSE subjected to laparoscopic ovariectomy in healthy mares did not demonstrate high difficulties in its execution nor post-surgical complications, confirming the practicality of the two techniques in the equine species.

A laparoscopia é técnica cirúrgica minimamente invasiva na qual se utiliza um laparoscópio inserido por via transabdominal por onde é possível inspecionar visualmente a superfície das vísceras e do peritônio e realizar procedimentos cirúrgicos. Até o momento, a laparoscopia diagnóstica em equinos em estação através da fossa paralombar exige que sejam realizados dois procedimentos (um de cada lado do animal) para visualização significativa da cavidade peritoneal, além do uso de laparoscópio específico. Com o intuito de desenvolver a laparoscopia diagnóstica em equinos em estação por acesso único, foi desenvolvida uma cânula óptica com espiral sem trocarte de 60 cm de comprimento que tem o objetivo de promover maior segurança ao acesso laparoscópico (penetração radial sem trocarte e sob visualização) e sustentação ao endoscópio flexível para que seja possível realizar exploração significativa de ambos os hemisférios abdominais dorsais por apenas uma das fossas paralombares. Foram utilizados seis equinos adultos, sendo três machos e três fêmeas, submetidos a jejum alimentar de 24 a 30 horas. Amostras de sangue para realização de hemograma e dosagem de fibrinogênio foram obtidas no D-Pré, D4 e D7. Os equinos foram sedados com cloridrato de detomidina, seguido de bloqueio anestésico local com cloridrato de lidocaína infundida na musculatura e tecido subcutâneo do local de incisão no centro da fossa paralombar. O procedimento cirúrgico teve início com incisão cutânea de cerca de 2 cm de comprimento. As camadas musculares da região foram afastadas através da rotação da cânula em sentido horário e o afastamento dos tecidos foi acompanhado pelas imagens produzidas pelo endoscópio flexível e reproduzidas no monitor. Logo da perfuração do peritônio, foi iniciada a exploração da cavidade. Após a exploração do lado ipsilateral ao local de acesso, foi realizada a transposição do conjunto cânula/endoscópio ventralmente à porção caudal do cólon descendente seguida de exploração do lado contralateral. Terminado o exame, o conjunto foi removido através da rotação da cânula no sentido anti-horário e a incisão de pele foi suturada junto ao subcutâneo. A movimentação da cânula entre os órgãos abdominais mostrou-se segura, no entanto, em dois animais houve lesão iatrogênica ao baço, evidenciada por discreta hemorragia local. Foi possível a identificação da maioria das estruturas abdominais descritas na literatura por procedimento laparoscópico em equinos em estação pelo acesso único. Em cinco dos seis animais houve formação de enfisema subcutâneo adjacente às feridas cirúrgicas. Os parâmetros hematológicos dos animais mantiveram-se dentro dos limites estabelecidos para a espécie, com exceção dos leucócitos totais, neutrófilos segmentados e fibrinogênio que demonstraram pequenas variações. A abordagem proposta mostrou-se viável e tem o potencial de tornar o procedimento mais rápido, prático e seguro e menos invasivo e oneroso, fomentando a aplicação da laparoscopia diagnóstica na espécie equina. / Laparoscopy is a minimally invasive surgical technique by using a laparoscope inserted through the abdominal wall where it is visually possible to inspect visceral surfaces and peritoneum, and to perform surgical procedures. Up to now, diagnostic laparoscopy in standing horses through the paralumbar fossa requires two procedures to be performed (one on each side of the animal) in order to significantly see the peritoneal cavity, in addition of using specific laparoscope. With the purpose to develop a diagnostic laparoscopy in standing horses through a single access, a trocarless threaded visual cannula 60 cm length was designed with the aim to promote higher safety to laparoscopic access (radial penetration without trocar and under visualization), and support to a flexible endoscope so that a significant exploration of both dorsal abdominal hemispheres through only one of the paralumbar fossa is possible to perform. Six mature horses were used: three males and three females. Food was withheld for 24 to 30 hours before surgery. Blood samples to perform complete blood count and fibrinogen metering were obtained in D-Pre, D4, and D7. The horses were sedated with detomidine hydrochloride, followed by local anesthesia blockade with lidocaine hydrochloride infused in the muscle and subcutaneous tissue at the incision place in the center of the paralumbar fossa. The surgical procedure started with a skin incision of nearly 2 cm length. Muscle layers of the region were moved away through clockwise cannula rotation and the distancing of the tissues were followed by images produced by the flexible endoscope and played on the monitor. After the peritoneum perforation, the cavity exploration was started. After the ipsilateral side exploration to the access site, the transposing of the set cannula/endoscope was performed ventrally to the rear portion of the descending colon followed by the contralateral side exploration. When the examination was finished, the set was removed through counterclockwise cannula rotation and the skin incision was stitched to the subcutaneous. Moving the cannula among the abdominal organs was shown to be safe; however, spleen iatrogenic injury in two animals was highlighted by minor local hemorrhage. The majority of the abdominal structures described in the literature was possible to identify through the laparoscopic procedure in standing horses by single access. There was subcutaneous emphysema formation adjoining the surgical wounds in five of the six animals. The hematological parameters of the animals were kept in the limits established for the species, with exception of total leucocytes, segmented neutrophils and fibrinogen which showed small variations. The proposed approach was shown to be feasible and has the potential to make the procedure faster, more practical and safe and less invasive and expensive, fostering the diagnostic laparoscopy application in the equine species.