Artificial Intelligence in Neurosurgery
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Rīgas Stradiņa universitāte
Rīga Stradiņš University
Rīga Stradiņš University
Abstract
KOPSAVILKUMS
Ievads:
Mākslīgā intelekta (AI) izmantošana var palīdzēt diagnosticēt un ārstēt mugurkaula slimības.
Metodes:
3 datubāzēs tika meklēta literatūra par mākslīgā intelekta izmantošanu mugurkaula ķirurģijā. Pētījumi tika iedalīti pirmsoperācijas, intraoperācijas un pēcoperācijas pētījumos. Atzinumi tika apkopoti, izmantojot nejaušo efektu kopsavilkuma darbības uztveršanas raksturlīknes.
Rezultāti:
Analizētajos 64 pētījumos pirmsoperācijas, intraoperācijas un pēcoperācijas aprūpei tika konstatēta attiecīgi 44 %, 11 % un 13 % metapieredze. Pirmsoperācijas gadījumā AI modeļu apvienotais jutīgums bija 80,1 % (95 % IK - 73,1 % līdz 85,6 %) un apvienotais specifiskums - 80,4 % (95 % IK - 72,8 % līdz 86,3 %). Intraoperācijas laikā AI modeļu apvienotais jutīgums bija 84,2 % (95 % KI - 73,2 % līdz 91,3 %) un apvienotais specifiskums bija 97,8 % (95 % KI - 90 % līdz 99,5 %). Pēcoperācijas gadījumā AI modeļu apvienotais jutīgums bija 77 % (95 % KI - 64,7 % līdz 86 %) un apvienotais specifiskums bija 92,5 % (95 % KI - 85,8 % līdz 96,2 %).
Secinājumi:
Lai gan AI ir lielāks potenciāls mugurkaula ķirurģijā, joprojām pastāv bažas par drošību un pierādījumu kvalitāti.
Atslēgas vārdi:
mugurkaula operācijas, mākslīgais intelekts, mākslīgais intelekts, jutīgums, specifiskums
ABSTRACT Introduction: The use of Artificial Intelligence (AI) can assist in diagnosing and treatment of spinal diseases. Methods: 3 databases were searched for literature on the application of AI in the spine surgery. The studies were divided into preoperative, intraoperative, and postoperative types. The findings were summarized using random-effects summary receiver operating characteristic curve. Results: In the 64 analyzed studies, a meta-prevalence of 44%, 11% and 13% was observed for preoperative, intraoperative, and postoperative care, respectively. For, preoperative, the pooled sensitivity of AI models was 80.1% (95% CI – 73.1% to 85.6%) and pooled specificity was 80.4% (95% CI – 72.8% to 86.3%). For, intraoperative, the pooled sensitivity of AI models was 84.2% (95% CI – 73.2% to 91.3%) and pooled specificity was 97.8% (95% CI – 90% to 99.5%). For postoperative, the pooled sensitivity of AI models was 77% (95% CI – 64.7% to 86%) and pooled specificity was 92.5% (95% CI – 85.8% to 96.2%). Conclusions: Although AI has increased potential in spine surgery, concerns remain regarding safety and quality of evidence. Keywords: Spinal surgeries, artificial intelligence, AI, sensitivity, specificity
ABSTRACT Introduction: The use of Artificial Intelligence (AI) can assist in diagnosing and treatment of spinal diseases. Methods: 3 databases were searched for literature on the application of AI in the spine surgery. The studies were divided into preoperative, intraoperative, and postoperative types. The findings were summarized using random-effects summary receiver operating characteristic curve. Results: In the 64 analyzed studies, a meta-prevalence of 44%, 11% and 13% was observed for preoperative, intraoperative, and postoperative care, respectively. For, preoperative, the pooled sensitivity of AI models was 80.1% (95% CI – 73.1% to 85.6%) and pooled specificity was 80.4% (95% CI – 72.8% to 86.3%). For, intraoperative, the pooled sensitivity of AI models was 84.2% (95% CI – 73.2% to 91.3%) and pooled specificity was 97.8% (95% CI – 90% to 99.5%). For postoperative, the pooled sensitivity of AI models was 77% (95% CI – 64.7% to 86%) and pooled specificity was 92.5% (95% CI – 85.8% to 96.2%). Conclusions: Although AI has increased potential in spine surgery, concerns remain regarding safety and quality of evidence. Keywords: Spinal surgeries, artificial intelligence, AI, sensitivity, specificity
Description
Medicīna
Medicine
Veselības aprūpe
Health Care
Medicine
Veselības aprūpe
Health Care
Keywords
mugurkaula operācijas, mākslīgais intelekts, mākslīgais intelekts, jutīgums, specifiskums, Spinal surgeries, artificial intelligence, AI, sensitivity, specificity