Joseph Jacob | Satsuma

Exploring Fairness and Performance Drivers Across State-of-the-Art Pulmonary Nodule Detection Algorithms

Mon, 04 May 2026 00:00:00 +0000

Researchers from Satsuma Lab have evaluated the fairness of deep learning based computer aided detection (CADe) systems for lung nodule detection in screening CT scans. Using data from a SUMMIT (London-based cohort), the study found that model performance remains consistent across sex and ethnic groups, despite imbalances in the training data. Results suggest that detection accuracy is driven more by nodule characteristics than by demographic factors. These findings support the equitable deployment of AI tools in future UK lung cancer screening programmes.

Further details can be found in John McCabe et al., 2025.

Expanded detection of early fibrotic phenotypes using lobar traction bronchiolectasis in lung cancer screening

Tue, 10 Mar 2026 00:00:00 +0000

A computationally frugal, open-source chest CT foundation model for thoracic disease detection in lung cancer screening programmes

Wed, 04 Feb 2026 00:00:00 +0000

Exploring Fairness and Performance Drivers Across State-of-the-Art Pulmonary Nodule Detection Algorithms

Sun, 21 Dec 2025 00:00:00 +0000

Predicting histopathological features of aggressiveness in lung cancer using CT radiomics: a systematic review

Sun, 01 Sep 2024 00:00:00 +0000

Enhancing cancer prediction in challenging screen-detected incident lung nodules using time-series deep learning

Mon, 03 Jun 2024 00:00:00 +0000

Evaluation of automated airway morphological quantification for assessing fibrosing lung disease

Sun, 31 Mar 2024 00:00:00 +0000

Optimising Chest X-Rays for Image Analysis by Identifying and Removing Confounding Factors

Mon, 20 Nov 2023 00:00:00 +0000

CenTime: Event-conditional modelling of censoring in survival analysis

Sun, 29 Oct 2023 00:00:00 +0000

A hybrid CNN-RNN approach for survival analysis in a Lung Cancer Screening study

Fri, 29 Sep 2023 00:00:00 +0000

Researchers from Satsuma Lab have introduced a hybrid approach combining Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) to predict long-term survival in a Lung Cancer Screening (LCS) study. It was demonstrated that incorporating the patient’s imaging follow-up history can lead to improvement in survival prediction. Delineating subjects at increased risk of cardiorespiratory mortality can alert clinicians to request further more detailed functional or imaging studies to improve the assessment of cardiorespiratory disease burden. Such strategies may uncover unsuspected and under-recognised pathologies thereby potentially reducing patient morbidity.

Further details can be found in Lu et al., 2023.

A hybrid CNN-RNN approach for survival analysis in a Lung Cancer Screening study

Thu, 03 Aug 2023 00:00:00 +0000

Automated airway quantification associates with mortality in idiopathic pulmonary fibrosis

Sat, 01 Jul 2023 00:00:00 +0000

Delineating associations of progressive pleuroparenchymal fibroelastosis in patients with pulmonary fibrosis

Thu, 11 May 2023 00:00:00 +0000

MisMatch: Calibrated Segmentation via Consistency on Differential Morphological Feature Perturbations with Limited Labels

Wed, 10 May 2023 00:00:00 +0000

Moucheng Named Finalist for MICCAI Young Scientist Award (Best Paper Award)

Wed, 10 May 2023 00:00:00 +0000

Satsuma Lab is delighted to announce that Moucheng was named a finalist for the MICCAI Young Scientist Award for his paper: Bayesian Pseudo Labels: Expectation Maximization for Robust and Efficient Semi-Supervised Segmentation (Semi-Supervised Segmentation with Pseudo Labels). A journal extension can be found here: Expectation Maximization Pseudo Labelling for Segmentation with Limited Annotations

The MICCAI Young Scientist Award recognizes the best papers that are first-authored by young scientists at the main MICCAI conference. This award is regarded as one of the most prestigious and most competitive award in the field of medical image computing. Each year has 5 winners, 15 finalists and 30 nominations. In 2022, there were 1825 total submissions, making Moucheng’s paper top 0.8 % among all of the submissions.

Utilisation of deep learning for COVID-19 diagnosis

Sat, 11 Feb 2023 00:00:00 +0000

Airway measurement by refinement of synthetic images improves mortality prediction in idiopathic pulmonary fibrosis

Sat, 08 Oct 2022 00:00:00 +0000

Bayesian Pseudo Labels: Expectation Maximization for Robust and Efficient Semi Supervised Segmentation

Tue, 27 Sep 2022 00:00:00 +0000

Deep Learning Based Long Term Mortality Prediction in the National Lung Screening Trial

Thu, 24 Mar 2022 00:00:00 +0000

Radiology of Bronchiectasis

Tue, 01 Mar 2022 21:30:42 +0000

Survival Analysis for Idiopathic Pulmonary Fibrosis using CT Images and Incomplete Clinical Data

Mon, 28 Feb 2022 00:00:00 +0000

Learning Morphological Feature Perturbation for Semi-Supervised Segmentation

Sun, 27 Feb 2022 00:00:00 +0000

Evaluation of automated airway morphological quantification for assessing fibrosing lung disease

Fri, 19 Nov 2021 00:00:00 +0000

A computationally efficient approach to segmentation of the aorta and coronary arteries using deep learning

Thu, 18 Feb 2021 00:00:00 +0000

Mortality in combined pulmonary fibrosis and emphysema patients is determined by the sum of pulmonary fibrosis and emphysema

Fri, 01 Jan 2021 00:00:00 +0000

Pleuroparenchymal fibroelastosis in idiopathic pulmonary fibrosis: Survival analysis using visual and computer-based computed tomography assessment

Fri, 01 Jan 2021 00:00:00 +0000

Disentangling Human Error from Ground Truth in Segmentation of Medical Images

Mon, 07 Dec 2020 00:00:00 +0000

Learning to Pay Attention to Mistakes

Sun, 06 Sep 2020 00:00:00 +0000