AI on H&E histology predicts invasive bladder cancer prognosis without genomics
⚖️ Transparency notice: this article was written with AI assistance and reviewed by the author, a medical oncologist.
Inteligencia artificial, LLMs y transformación digital aplicados al flujo de trabajo clínico.
⚖️ Transparency notice: this article was written with AI assistance and reviewed by the author, a medical oncologist.

IA predice estado MSI (microsatellite instability) en cáncer colorrectal desde TC e histología H&E, sin secuenciación — cuando la imagen habla de biología tumoral.

IA sobre histología H&E predice riesgo de recaída en cáncer de vejiga infiltrante sin genómica extra (feature extraction de whole-slide + score pronóstico).

La FDA propone 7 criterios (Congruence, Coverage, Constraint, Consistency, Comprehension, Compliance, Completeness) para evaluar datos médicos sintéticos.
ABUS + AI to classify BIRADS 3-4 lesions: can it cut unnecessary biopsies? University of Bari study on 68 lesions.

Título: Ecografía automatizada (ABUS) y IA: clasificar mejor las lesiones BIRADS 3-4 sin disparar biopsias innecesarias Tipo: A (IA imagen médica) Tags: ecografía mamaria, ABUS, inteligencia artificial, BIRADS, biopsia PMID: 42572099 DOI: 1…

Cáncer ginecológico avanzado: inmunoterapia y ADCs ya no son experimentales. La selección por biomarcador (MMR, FRα, PD-L1) ya es el presente.

Tucatinib + T-DXd in advanced HER2+: the TKI + ADC combo achieves 51% response rate and 28 months overall survival in heavily pretreated patients (HER2CLIMB-04).

T cell engagers and antibody-drug conjugates are the 'guided missiles' that come after checkpoint inhibitors. They are already used in oncology and are beginning to cross into autoimmune disease.

Radiotherapy does more than kill tumor cells: it shapes the immune microenvironment. Novel delivery modalities (SBRT, LATTICE, LDRT) may boost immunotherapy depending on how the dose is distributed.