RESUMEN
Frozen sections are a useful pathologic tool, but variable image quality may impede the use of artificial intelligence and machine learning in their interpretation. We aimed to identify the current research on machine learning models trained or tested on frozen section images. We searched PubMed and Web of Science for articles presenting new machine learning models published in any year. Eighteen papers met all inclusion criteria. All papers presented at least one novel model trained or tested on frozen section images. Overall, convolutional neural networks tended to have the best performance. When physicians were able to view the output of the model, they tended to perform better than either the model or physicians alone at the tested task. Models trained on frozen sections performed well when tested on other slide preparations, but models trained on only formalin-fixed tissue performed significantly worse across other modalities. This suggests not only that machine learning can be applied to frozen section image processing, but also use of frozen section images may increase model generalizability. Additionally, expert physicians working in concert with artificial intelligence may be the future of frozen section histopathology.
Asunto(s)
Inteligencia Artificial , Secciones por Congelación , Humanos , Redes Neurales de la Computación , Aprendizaje AutomáticoAsunto(s)
Carcinoma Ductal de Mama/terapia , Interleucina-2/administración & dosificación , Neoplasias Cutáneas/terapia , Neoplasias de la Mama Triple Negativas/patología , Adulto , Carcinoma Ductal de Mama/secundario , Femenino , Humanos , Inyecciones Intralesiones , Neoplasias Cutáneas/secundario , Resultado del Tratamiento , Neoplasias de la Mama Triple Negativas/terapiaRESUMEN
Pregnancy is a conjectured risk factor for pyoderma gangrenosum (PG), an autoinflammatory neutrophilic dermatosis characterized by painful ulcers. Even so, there are no available treatment guidelines for those with PG who are pregnant or breastfeeding. To describe existing treatment options, we systematically reviewed the literature on PG treatment in pregnant or breastfeeding patients. A search over four databases was completed in October 2022. Independent reviewers accomplished screening and data extraction. 18 articles met the inclusion criteria. 15 cases involved the treatment of PG during pregnancy, and three cases involved the treatment of PG while breastfeeding. Most patients did not have a history of PG prior to pregnancy (77.7%), and most did not have PG-associated comorbidity (61.1%). Of the cases involving treatment of PG during pregnancy, the majority (73%) found treatment success with a systemic corticosteroid (SCS). Only three cases reported an adverse outcome, including premature rupture of membranes and premature birth (16.7%); all these cases involved treatment with a SCS at >0.5 mg/kg/day during pregnancy. We present a treatment algorithm for pregnant or breastfeeding patients with PG. Our findings suggest prioritizing topicals and TNF inhibitors due to more favorable side effect profiles. However, there is a paucity of data on the safety of PG therapies in pregnancy and breastfeeding, and thus, controlled studies and pregnancy registries must be pursued.
RESUMEN
With the recent unprecedented shift toward contactless healthcare solutions, providers should recall the value that proper palpation adds to dermatologic practice. We present a case that demonstrates the limitations of touchless care and how proper palpation during skin cancer examinations may impact cosmetic outcomes. Our patient is an 86-year-old male patient with Sezary syndrome and monoclonal B-cell lymphocytosis whose squamous cell carcinoma invasion was missed by visual inspection alone. He delayed treatment of his biopsy-proven squamous cell carcinoma for 15 months. On follow-up, visual examination only showed a well-healed biopsy scar, and treatment was delayed another 2 months. Finally, thorough physical examination found perineural invasion. This helped guide the Mohs approach, but due to the delays resulted in a larger final defect and poorer cosmetic outcomes. Proper, deep palpation of skin lesions, especially prior biopsy sites, is imperative to the treatment of skin cancer in cosmetically sensitive areas. Biopsy scars on the face often heal well, and visual only inspection may miss crucial details. This case also reminds dermatologists of the importance of patient education in the prompt treatment of skin cancer for the best cosmetic outcomes.