Medical AI brings together our reporting on machine learning for clinical and biomedical problems, from diagnosis and prognosis to medical image analysis and decision support. Because mistakes in this domain carry a human cost, we pay close attention to evaluation, calibration, uncertainty, and the gap between benchmark numbers and bedside reliability. Expect grounded explainers of recent research rather than uncritical product announcements.
Ask anyone who has tried to build a medical image segmentation model what the hardest part is, and the honest answer is rarely the neural network architecture. It is getting enough labeled data. A useful segmentation dataset needs pixel accurate…
DFCPS: Teaching a Polyp Detector With Barely Any Labels Read More »
A radiologist in a busy teaching hospital reads maybe forty brain studies before lunch. Glioma, meningioma, pituitary adenoma, or nothing at all. Most of those calls are easy. A handful are not, and the hard ones tend to be hard…
What 99.62% Accuracy on Brain Tumor MRI Actually Means Read More »
A dermatologist looking at a mole under a dermatoscope is doing something a plain convolutional network struggles with. She is reading fine texture right at the border of the lesion while also holding the whole shape in mind at once,…
EG-VAN Explained, Dual Branch Attention for Skin Cancer Scans Read More »
A dermatologist bends over a dermoscope and studies a brown patch no bigger than a pencil eraser on a patient’s back. Is it an ordinary mole, an early melanoma, or a benign keratosis that only pretends to be dangerous? Human…
Skin Lesion Recognition With Global Local Attention Read More »
Skin cancer is common enough, and dangerous enough when caught late, that the stakes for this kind of research are genuinely high. Melanoma’s ten year survival rate falls below 39 percent when it is detected at an advanced stage, according…
A Top Scoring Skin Cancer AI Missed Most Basal Cell Cases Read More »
Skin cancer screening was one of the first places deep learning actually earned its hype. A 2017 Nature paper by Esteva and colleagues showed a convolutional network matching board certified dermatologists at distinguishing benign moles from malignant ones, and the…
Reversing Adversarial Attacks on Skin Cancer AI With Multiscale Noise Read More »
Lung cancer kills more people worldwide than any other cancer, and the reason is timing more than treatment. A lung nodule found early, while it is still small and localized, is a very different clinical situation from one found after…
Why LungCT-NET Stacks Four Networks Instead of Trusting One Read More »
The Segment Anything Model, usually shortened to SAM, arrived in 2023 as a general purpose segmentation system built on a vision transformer encoder, a prompt encoder and a lightweight mask decoder. It was trained on eleven million images carrying one…
SAM Was Never Built for Hospitals, So Researchers Made It Useful Anyway Read More »
Brain tumors do not follow a template. The same diagnosis, glioma for instance, can show up as wildly different shapes, sizes and locations from one patient’s scan to the next, while different tumor types can sometimes look deceptively similar to…
DEF-SwinE2NET Adds Two Small Modules to EfficientNetV2S and Nearly Erases Its Errors Read More »
Medical image segmentation sounds like a narrow technical task, draw a boundary around the diseased tissue, but the difficulty hides in what that boundary actually needs to capture. A tumor in a breast ultrasound image does not have a clean…
Why 3DL-Net’s Dendritic Neurons Only Help When Paired With Its Pyramid Module Read More »