Lung cancer has been a major cause of death among types of cancers in the world. In the early stages, lung nodules can be detected by the aid of imaging modalities such as Computed Tomography (CT). In this stage, radiologists look for irregular rounded-shaped nodules in the lung which are normally less than 3 centimeters in diameter. Recent advancements in image analysis have proven that images contain more information than regular parameters such as intensity, histogram and morphological details. Therefore, in this project we have focused on extracting quantitative, hand-crafted features from nearly 1400 lung CT images to train a variety of classifiers based on them. In the first experiment, in total 424 Radiomics features per image has been used to train classifiers such as: Random Forest (RF), Support Vector Machine (SVM), Decision Tree (DT), Naive Bayes (NB), Linear Discriminant Analysis (LDA) and Multi-Layer Perceptron (MLP). In the second experiment, we evaluate each feature category separately with our classifiers. The third experiment includes wrapper feature selection methods (Forward/Backward/Recursive) and filter-based feature selection methods (Fisher score, Gini Index and Mutual information). They have been implemented to find the most relevant feature set in model construction. Performance of each learning method has been evaluated by accuracy score, wherewe achieved the highest accuracy of 78% with Random Forest classifier (74% in 5-fold average) and 0.82 Area Under the Receiver Operating Characteristics (AUROC) curve. After RF, NB and MLP showed the best average accuracy of 71.4% and 71% respectively.

Os tumores carcinoides típicos broncopulmonares (TCTB) são considerados neoplasias bem diferenciadas e as menos agressivas dentro do espectro dos tumores neuroendócrinos. Entretanto, metástases linfonodais e hematogênicas tem sido encontradas em número considerável de casos e não existem, até o momento, estudos relacionados aos mecanismos de escape imune tumoral em TCTB. Alguns trabalhos tem relacionado a expressão da enzima indoleamine 2,3-dioxigenase (IDO) em células neoplásicas como fator responsável pela aquisição de tolerância tumoral. Além disso, os níveis de infiltração linfocitária intratumoral parecem estar associados com prognóstico e sobrevida nesses tumores. O principal objetivo deste estudo foi determinar os níveis de expressão intratumorais da IDO e sua possível aplicação como marcador biológico de metástases em TCTB. Além disso, também foi estabelecido o padrão de infiltração linfocitária intratumoral e analisada sua provável correlação com os níveis de expressão da IDO. Portanto, realizou-se uma coorte retrospectiva multicêntrica no qual 64 pacientes submetidos à cirurgia de ressecção de TCTB entre 1981 e 2003 foram selecionados. O período de seguimento pós-operatório foi de 5 anos e a ocorrência de metástases linfonodais (hilar ou mediastinal) e hematogênicas foi avaliada através de tomografia computadorizada. Os níveis da expressão da IDO e de infiltração linfocitária intratumoral foram avaliados através de estudo imunohistoquímico. Os resultados obtidos mostraram que dos 64 pacientes selecionados, 17 (26.5%) apresentaram qualquer tipo de metátases durante o estudo: linfonodal, hematogênica ou ambas. A expressão da IDO foi encontrada em níveis diferentes de intensidade em mais de 80% das células dos TBCT. Entretanto, a análise univariada não mostrou nenhuma diferença significante na expressão da IDO entre grupos com e sem metástase (p=0,9 e p=0,3 pela análise semi-quantitativa e quantitativa, respectivamente). A quantificação dos linfócitos em todos os grupos estudados demonstrou predominância de linfócitos T CD8+, quando comparado aos linfócitos T CD4+(p< 0.01). Nenhuma diferença na infiltração intratumoral de linfócitos T CD8+ foi encontrada entre grupos com e sem metástase (p=0,98). Entretanto, a quantificação de linfócitos T CD4+ foi nula nos grupos com qualquer tipo de metástase (p=0,01), e nos casos com metástase linfonodal (p=0,02). Nenhuma correlação entre os níveis da expressão da IDO e da infiltração linfocitária intratumoral foi identificada nos grupos analisados (r= -0.2 e p=0,1 para ambos os grupos). Conclui-se que, a expressão intratumoral da IDO não apresenta correlação com a ocorrência de metástase nos TBCT. Apesar de nenhuma diferença ter sido identificada na infiltração intratumoral de linfócitos T CD8+ nos grupos com e sem metástase, a ausência de infiltração de linfócitos T CD4+ está associado à ocorrência do evento estudado. Estes linfócitos parecem conferir um efeito protetor evitando o escape tumoral / Typical bronchopulmonary carcinoid tumors (TBCT) are considered the less aggressive neoplasm within the spectrum of neuroendocrine tumors. However, regional nodes and haematogenic metastasis occur in a considerable rate and no data regarding immune escape mechanisms in these tumors are available. Some studies have implicated indoleamine 2,3 dioxygenase (IDO) expression in malignant cells as the responsible for tumor tolerance. Also, levels of tumor infiltrating lymphocytes (TILs) seem to be related with prognosis and survival. Our aim in this study was to determine intratumoral IDO expression levels and the value of this variable as a predictive marker of TBCT metastasis. Thus, TILs pattern was determined and correlation with intratumoral IDO expression analyzed. For this purpose, a multicenter retrospective cohort study was performed and 64 patients operated on for TBCT between 1981 and 2003 were enrolled. Follow-up period was 5 years and regional or haematogenic metastasis was assessed by computerized tomography (CT) scan. Levels of IDO expression and TILs were assessed by immunohistochemical study. The results obtained showed that of all 64 patients, 17 (26,5%) presented with any metastasis during the study: regional nodes, haematogenic or both. IDO expression was found in different intensity levels in over 80% of TBCT cells. However, univariate analysis showed no significant difference in IDO expression between groups with and without metastasis (p=0,9 and p=0,3 for semi-quantitative and quantitative analysis respectively). TILs quantification in all studied groups demonstrated predominance of CD8+ TILs when compared to CD4+ TILs (p<0,01). No difference in CD8+ TILs was found between groups with and without metastasis (p=0,98). However CD4+ TILs quantification was null in the groups with any metastasis (p=0,01), and regional nodes metastasis (p=0,02). No correlation between IDO expression levels and TILs was identified in all analyzed groups(r= -0,2 and p=0,1 for both groups). In conclusion, these data shows that intratumoral IDO expression do not correlate with TBCT metastasis. Even though no difference in CD8+ TILs between groups with and without metastasis was found, absence of CD4+ TILs is associated with the studied event. These cells seem to confer a protective effect against tumoral immune escape

Os tumores carcinoides típicos broncopulmonares (TCTB) são considerados neoplasias bem diferenciadas e as menos agressivas dentro do espectro dos tumores neuroendócrinos. Entretanto, metástases linfonodais e hematogênicas tem sido encontradas em número considerável de casos e não existem, até o momento, estudos relacionados aos mecanismos de escape imune tumoral em TCTB. Alguns trabalhos tem relacionado a expressão da enzima indoleamine 2,3-dioxigenase (IDO) em células neoplásicas como fator responsável pela aquisição de tolerância tumoral. Além disso, os níveis de infiltração linfocitária intratumoral parecem estar associados com prognóstico e sobrevida nesses tumores. O principal objetivo deste estudo foi determinar os níveis de expressão intratumorais da IDO e sua possível aplicação como marcador biológico de metástases em TCTB. Além disso, também foi estabelecido o padrão de infiltração linfocitária intratumoral e analisada sua provável correlação com os níveis de expressão da IDO. Portanto, realizou-se uma coorte retrospectiva multicêntrica no qual 64 pacientes submetidos à cirurgia de ressecção de TCTB entre 1981 e 2003 foram selecionados. O período de seguimento pós-operatório foi de 5 anos e a ocorrência de metástases linfonodais (hilar ou mediastinal) e hematogênicas foi avaliada através de tomografia computadorizada. Os níveis da expressão da IDO e de infiltração linfocitária intratumoral foram avaliados através de estudo imunohistoquímico. Os resultados obtidos mostraram que dos 64 pacientes selecionados, 17 (26.5%) apresentaram qualquer tipo de metátases durante o estudo: linfonodal, hematogênica ou ambas. A expressão da IDO foi encontrada em níveis diferentes de intensidade em mais de 80% das células dos TBCT. Entretanto, a análise univariada não mostrou nenhuma diferença significante na expressão da IDO entre grupos com e sem metástase (p=0,9 e p=0,3 pela análise semi-quantitativa e quantitativa, respectivamente). A quantificação dos linfócitos em todos os grupos estudados demonstrou predominância de linfócitos T CD8+, quando comparado aos linfócitos T CD4+(p< 0.01). Nenhuma diferença na infiltração intratumoral de linfócitos T CD8+ foi encontrada entre grupos com e sem metástase (p=0,98). Entretanto, a quantificação de linfócitos T CD4+ foi nula nos grupos com qualquer tipo de metástase (p=0,01), e nos casos com metástase linfonodal (p=0,02). Nenhuma correlação entre os níveis da expressão da IDO e da infiltração linfocitária intratumoral foi identificada nos grupos analisados (r= -0.2 e p=0,1 para ambos os grupos). Conclui-se que, a expressão intratumoral da IDO não apresenta correlação com a ocorrência de metástase nos TBCT. Apesar de nenhuma diferença ter sido identificada na infiltração intratumoral de linfócitos T CD8+ nos grupos com e sem metástase, a ausência de infiltração de linfócitos T CD4+ está associado à ocorrência do evento estudado. Estes linfócitos parecem conferir um efeito protetor evitando o escape tumoral / Typical bronchopulmonary carcinoid tumors (TBCT) are considered the less aggressive neoplasm within the spectrum of neuroendocrine tumors. However, regional nodes and haematogenic metastasis occur in a considerable rate and no data regarding immune escape mechanisms in these tumors are available. Some studies have implicated indoleamine 2,3 dioxygenase (IDO) expression in malignant cells as the responsible for tumor tolerance. Also, levels of tumor infiltrating lymphocytes (TILs) seem to be related with prognosis and survival. Our aim in this study was to determine intratumoral IDO expression levels and the value of this variable as a predictive marker of TBCT metastasis. Thus, TILs pattern was determined and correlation with intratumoral IDO expression analyzed. For this purpose, a multicenter retrospective cohort study was performed and 64 patients operated on for TBCT between 1981 and 2003 were enrolled. Follow-up period was 5 years and regional or haematogenic metastasis was assessed by computerized tomography (CT) scan. Levels of IDO expression and TILs were assessed by immunohistochemical study. The results obtained showed that of all 64 patients, 17 (26,5%) presented with any metastasis during the study: regional nodes, haematogenic or both. IDO expression was found in different intensity levels in over 80% of TBCT cells. However, univariate analysis showed no significant difference in IDO expression between groups with and without metastasis (p=0,9 and p=0,3 for semi-quantitative and quantitative analysis respectively). TILs quantification in all studied groups demonstrated predominance of CD8+ TILs when compared to CD4+ TILs (p<0,01). No difference in CD8+ TILs was found between groups with and without metastasis (p=0,98). However CD4+ TILs quantification was null in the groups with any metastasis (p=0,01), and regional nodes metastasis (p=0,02). No correlation between IDO expression levels and TILs was identified in all analyzed groups(r= -0,2 and p=0,1 for both groups). In conclusion, these data shows that intratumoral IDO expression do not correlate with TBCT metastasis. Even though no difference in CD8+ TILs between groups with and without metastasis was found, absence of CD4+ TILs is associated with the studied event. These cells seem to confer a protective effect against tumoral immune escape

Medical imagery is increasingly evolving towards higher resolution and throughput. The increasing volume of data and the usage of multiple and often novel imaging modalities necessitates the use of mathematical and computational techniques for quicker, more accurate and more robust analysis of medical imagery. The fields of computer vision and machine learning provide a rich set of techniques that are useful in medical image analysis, in tasks ranging from segmentation to classification and population analysis, notably by integrating the qualitative knowledge of experts in anatomy and the pathologies of various disorders and making it applicable to the analysis of medical imagery going forward. The object of the proposed research is exactly to explore various computer vision and machine learning methods with a view to the improved analysis of multiple modalities of brain and cardiac imagery, towards enabling the clinical goals of studying schizophrenia, brain tumors (meningiomas and gliomas in particular) and cardiovascular disorders. In the first project, a framework is proposed for the segmentation of tubular, branched anatomical structures. The framework uses the tubular surface model which yields computational advantages and further incorporates a novel automatic branch detection algorithm. It is successfully applied to the segmentation of neural fiber bundles and blood vessels. In the second project, a novel population analysis framework is built using the shape model proposed as part of the first project. This framework is applied to the analysis of neural fiber bundles towards the detection and understanding of schizophrenia. In the third and final project, the use of mass spectrometry imaging for the analysis of brain tumors is motivated on two fronts, towards the offline classification analysis of the data, as well as the end application of intraoperative detection of tumor boundaries. SVMs are applied for the classification of gliomas into one of four subtypes towards application in building appropriate treatment plans, and multiple statistical measures are studied with a view to feature extraction (or biomarker detection). The problem of intraoperative tumor boundary detection is formulated as a detection of local minima of the spatial map of tumor cell concentration which in turn is modeled as a function of the mass spectra, via regression techniques.

Das erste gestellte Ziel war es, die histologischen Strukturen eines Gewebedünnschnittes anhand der aufgenommenen FT-IR-Spektren sichtbar zu machen und diese mit dem konventionell gefärbten Schnitt und der autoradiographischen Aufnahme zu vergleichen. Dazu wurde ein Messsystem bestehend aus einem FT-IR-Spektrometer mit Mikroskop und einem computergesteuerten XY-Tisch aufgebaut und die notwendige Steuer- und Auswerte-Software entwickelt. Es konnte gezeigt werden, dass sich die FT-IR-Spektren mit geeigneten Auswerteverfahren zur Darstellung der histologischen Strukturen nutzen lassen. Dazu wurden zwei verschiedene Methoden eingesetzt, die PCA und die Fuzzy-Clusterung (FCM). Im zweiten Teil dieser Arbeit sollte ein Klassifikations-Algorithmus gefunden werden, mit dessen Hilfe sich Spektren von unbekannten Gewebeproben vorher definierten Modellen zuordnen lassen. Dazu wurde eine Spektren-Datenbank aus mehr als einhundert Gewebeproben angelegt. Aus dieser Datenbank wurden einige zehntausend Spektren ausgewählt und zu Modell-Datensätzen für sechs verschiedene Gewebetypen zusammengefasst. Für die Zuordnung unbekannter Spektren zu diesen Modellen wurde ein SIMCA-Klassifikations-Algorithmus entwickelt sowie ein LDA-Algorithmus eingesetzt. Für beide Methoden wurde die Klassi-fikations-Leistung anhand der Spezifität und Sensitivität bestimmt. Beide Klassifikations-Algorithmen führten zu guten Ergebnissen. Der SIMCA-Algorithmus erreichte eine Spezifität zwischen 97 % und 100 %, sowie eine Sensitivität zwischen 62 % und 78 % (bei einem Vertrauensintervall von 97,5 %). Der LDA-Algorithmus ermöglichte eine etwas bessere Sensitivität von 72 % bis 90 %, auf Kosten der Spezifität, welche zwischen 90 % und 98 % lag. Zusammenfassend kann festgestellt werden, dass sich die FT-IR-Mikrospektroskopie und die vorgestellten Auswerte-Algorithmen sehr gut zur Klassifizierung von Gewebedünnschnitten eignen.

This study explores the potential of artificial intelligence (AI) in enhancing the diagnosis of brain tumours, specifically through a comparative analysis of two advanced deep learning (DL) models, ResNet50 and YOLOv8, applied to detect and classify brain tumours in MRI images. The study addresses the critical need for rapid and accurate diagnostic tools in the medical field, given the complexity and diversity of brain tumours. The research was motivated by the potential benefits AI could offer to medical diagnostics, particularly in terms of speed and accuracy, which are crucial for effective patient treatment and outcomes. The performance of the ResNet50 and YOLOv8 models was evaluated on a dataset of 7023 MRI images across four tumour types. Key metrics used were accuracy, precision, recall, specificity, F1-score, and processing time, to identify which model performs better in detecting and classifying brain tumours. The findings demonstrates that although both models exhibit high performance, YOLOv8 surpasses ResNet50 in most metrics, particularly showing advantages in speed. The findings highlight the effectiveness advanced DL models in medical image analysis, providing a significant advancement in brain tumour diagnosis. By offering a thorough comparative analysis of two commonly used DL models, aligning with ongoing approaches to integrate AI into practical medical application, and highlighting their potential uses, this study advances the area of medical AI providing insight into the knowledge required for the deployment of future AI diagnostic tools.