Multimodal system for skin cancer detection

Melanoma detection is vital for early diagnosis and effective treatment. While deep learning models on dermoscopic images have shown promise, they require specialized equipment, limiting their use in broader clinical settings. This study introduces a multi-modal melanoma detection system using conventional photo images, making it more accessible and versatile. Our system integrates image data with tabular metadata, such as patient demographics and lesion characteristics, to improve detection accuracy. It employs a multi-modal neural network combining image and metadata processing and supports a two-step model for cases with or without metadata. A three-stage pipeline further refines predictions by boosting algorithms and enhancing performance. To address the challenges of a highly imbalanced dataset, specific techniques were implemented to ensure robust training. An ablation study evaluated recent vision architectures, boosting algorithms, and loss functions, achieving a peak Partial ROC AUC of 0.18068 (0.2 maximum) and top-15 retrieval sensitivity of 0.78371. Results demonstrate that integrating photo images with metadata in a structured, multi-stage pipeline yields significant performance improvements. This system advances melanoma detection by providing a scalable, equipment-independent solution suitable for diverse healthcare environments, bridging the gap between specialized and general clinical practices.

Interactive decision support system for lung cancer segmentation

This paper studies Clinical Intelligent Decision Support Systems (CIDSSs) for lung cancer segmentation, which are based on deep neural nets. A new interactive CIDSS is proposed and compared with previous approaches. Addition-ally, the purpose uncertainty problem in building interactive systems is discussed, and criteria for measuring both quality and amount of user feedback are proposed. In order to automate system evaluation, a new algorithm was used to simulate expert feedback. The proposed interactive CIDSS outperforms previous approaches (both interactive and noninteractive) on the task of lung lesion segmentation. This ap-proach looks promising both in terms of quality and expert user experience. At the same time, this paper discusses a bunch of possible modifications that can be done to improve both evaluation criteria and proposed CIDSS in future works.

Semi-Supervised Segmentation of Functional Tissue Units at the Cellular Level

We present a new method for functional tissue unit segmentation at the cellular level, which utilizes the latest deep learning semantic segmentation approaches together with domain adaptation and semi-supervised learning techniques. This approach allows for minimizing the domain gap, class imbalance, and captures settings influence between HPA and HubMAP datasets. The presented approach achieves comparable with state-of-the-art-result in functional tissue unit segmentation at the cellular level. The source code is available at https://github.com/VSydorskyy/hubmap_2022_htt_solution