{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":45867,"databundleVersionId":6924515,"sourceType":"competition"}],"dockerImageVersionId":30626,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"- The goal of the UBC Ovarian Cancer subtypE clAssification and outlier detectioN (UBC-OCEAN) competition is to classify ovarian cancer subtypes. You will build a model trained on the world's most extensive ovarian cancer dataset of histopathology images obtained from more than 20 medical centers.\n\n\n# \n\n\nOvarian carcinoma, the deadliest cancer of the female reproductive system, encompasses a spectrum of distinct subtypes, each with unique characteristics. Accurate identification of these subtypes is crucial for tailoring effective treatment plans. However, the current reliance on pathologists for diagnosis poses challenges in terms of consistency and accessibility, especially in underserved communities. Leveraging data science offers a promising avenue to revolutionize ovarian cancer diagnosis and address these critical issues. This brief explores the potential of data-driven solutions in improving subtype identification and subsequently advancing personalized treatment strategies for ovarian carcinoma.","metadata":{}},{"cell_type":"markdown","source":"# \n\nThe dataset comprises images related to ovarian carcinoma, categorized into two main types: whole slide images (WSI) and tissue microarray (TMA). WSI images are captured at a 20x magnification, potentially yielding large file sizes. TMAs, on the other hand, are smaller in dimensions (approximately 4,000x4,000 pixels) but at a higher 40x magnification.\n\nIn the test set, images originate from different source hospitals compared to the training set. Notably, some of the test set's largest images are substantial, with dimensions nearing 100,000 x 50,000 pixels. It's essential to be prepared for diverse scenarios, including variations in image dimensions, quality, staining techniques, and more.\n\nThe test set consists of approximately 2,000 images, with the majority being TMAs. The overall dataset size is substantial, totaling 550 GB. Loading the data will require significant time and resources.\n\nPlease note that a few of the largest test set images may not entirely fit into memory on a notebook equipped with a GPU. A solution is under investigation, with updates expected around the week of October 18th.\n\n**CSV Files:**\nFor the train and test sets, accompanying CSV files provide crucial labels and information:\n\n**image_id:** A unique identifier for each image.\n**label:** The target class indicating subtypes of ovarian cancer, such as CC, EC, HGSC, LGSC, MC, or Other. Notably, the \"Other\" class is exclusive to the test set, highlighting the challenge of identifying outliers.\n**image_width:** The width of the image in pixels.\n**image_height:** The height of the image in pixels.\n**is_tma:** A binary value indicating whether the slide is a tissue microarray. This information is only available for the train set.\n\nAdditionally, the dataset includes a folder named [train/test]_thumbnails containing smaller .png versions of the whole slide images. Thumbnails, however, are not provided for TMAs.","metadata":{}},{"cell_type":"markdown","source":"# A brief descriptions of each subtype of ovarian carcinoma:
Exploratory Data Analysis (EDA)📊
Build the CNN Model and Prediction
\"Your positive feedback and upvotes are incredibly appreciated! They inspire me to create more valuable content and help others in their learning journey. Your support fosters a vibrant community of knowledge-sharing. Thank you for considering an upvote, and best wishes on your learning journey!\" 😊📌
","metadata":{}}]}