RESUMO
Techniques of artificial intelligence (AI) are increasingly used in the treatment of patients, such as providing a diagnosis in radiological imaging, improving workflow by triaging patients or providing an expert opinion based on clinical symptoms; however, such AI techniques also hold intrinsic risks as AI algorithms may point in the wrong direction and constitute a black box without explaining the reason for the decision-making process.This article outlines a case where an erroneous ChatGPT diagnosis, relied upon by the patient to evaluate symptoms, led to a significant treatment delay and a potentially life-threatening situation. With this case, we would like to point out the typical risks posed by the widespread application of AI tools not intended for medical decision-making.
Assuntos
Ataque Isquêmico Transitório , Humanos , Ataque Isquêmico Transitório/diagnóstico , Ataque Isquêmico Transitório/etiologia , Inteligência Artificial , Diagnóstico Tardio , Algoritmos , Tomada de Decisão ClínicaRESUMO
BACKGROUND: This study aimed to describe lesion-specific management of thoracic tumors referred for consideration of image-guided thermal ablation (IGTA) at a newly established multidisciplinary ablation conference. METHODS: This retrospective single-center cohort study included consecutive patients with non-small cell lung cancer (NSCLC) or thoracic metastases evaluated from June 2020 to January 2022 in a multidisciplinary conference. Outcomes included the management recommendation, treatments received (IGTA, surgical resection, stereotactic body radiation therapy [SBRT], multimodality management), and number of tumors treated per patient. Pearson's chi-square test was used to assess for a change in management, and Poisson regression was used to compare the number of tumors by treatment received. RESULTS: The study included 172 patients (58 % female; median age, 69 years; 56 % thoracic metastases; 27 % multifocal primary lung cancer; 59 % ECOG 0 [range, 0-3]) assessed in 206 evaluations. For the patients with NSCLC, IGTA was considered the most appropriate local therapy in 12 %, equal to SBRT in 22 %, and equal to lung resection in 3 % of evaluations. For the patients with thoracic metastases, IGTA was considered the most appropriate local therapy in 22 %, equal to SBRT in 12 %, and equal to lung resection in 3 % of evaluations. Although all patients were referred for consideration of IGTA, less than one third of patients with NSCLC or thoracic metastases underwent IGTA (p < 0.001). Multimodality management allowed for treatment of more tumors per patient than single-modality management (p < 0.01). CONCLUSIONS: Multidisciplinary evaluation of patients with thoracic tumors referred for consideration of IGTA significantly changed patient management and facilitated lesion-specific multimodality management.
Assuntos
Carcinoma Pulmonar de Células não Pequenas , Neoplasias Pulmonares , Radiocirurgia , Humanos , Feminino , Idoso , Masculino , Carcinoma Pulmonar de Células não Pequenas/patologia , Neoplasias Pulmonares/patologia , Estudos de Coortes , Estudos Retrospectivos , Resultado do TratamentoRESUMO
BACKGROUND. High-frequency jet ventilation (HFJV) facilitates accurate probe placement in percutaneous ablation of lung tumors but may increase risk for adverse events, including systemic air embolism. OBJECTIVE. The purpose of this study was to compare major adverse events and procedural efficiency of percutaneous lung ablation with HFJV under general anesthesia to spontaneous respiration (SR) under moderate sedation. METHODS. This retrospective study included consecutive adults who underwent CT-guided percutaneous cryoablation of one or more lung tumors with HFJV or SR between January 1, 2017, and May 31, 2023. We compared major adverse events (Common Terminology Criteria for Adverse Events grade ≥ 3) within 30 days postprocedure and hospital length of stay (HLOS) of 2 days or more using logistic regression analysis. We compared procedure time, room time, CT guidance acquisition time, CT guidance radiation dose, total radiation dose, and pneumothorax using generalized estimating equations. RESULTS. Overall, 139 patients (85 women, 54 men; median age, 68 years) with 310 lung tumors (82% metastases) underwent 208 cryoablations (HFJV, n = 129; SR, n = 79). HFJV showed greater rates than SR for the treatment of multiple tumors per session (43% vs 19%, respectively; p = .02) and tumors in a nonperipheral location (48% vs 24%, p < .001). Major adverse event rate was 8% for HFJV and 5% for SR (p = .46). No systemic air embolism occurred. HLOS was 2 days or more in 17% of sessions and did not differ significantly between HFJV and SR (p = .64), including after adjusting for probe number per session, chronic obstructive pulmonary disease, and operator experience (p = .53). Ventilation modalities showed no significant difference in procedure time, CT guidance acquisition time, CT guidance radiation dose, or total radiation dose (all p > .05). Room time was longer for HFJV than SR (median, 154 vs 127 minutes, p < .001). For HFJV, the median anesthesia time was 136 minutes. Ventilation modalities did not differ in the frequencies of pneumothorax or pneumothorax requiring chest tube placement (both p > .05). CONCLUSION. HFJV appears to be as safe as SR but had longer room times. HFJV can be used in complex cases without significantly impacting HLOS of 2 days or more, procedure time, or radiation exposure. CLINICAL IMPACT. Selection of the ventilation modality during percutaneous lung ablation should be based on patient characteristics and anticipated procedural requirements as well as operator preference.