• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) ¸¶Ãë¿ë óġ ÃßõÀ» À§ÇÑ ¼³¸í °¡´ÉÇÑ µö´º·² ³×Æ®¿öÅ©
¿µ¹®Á¦¸ñ(English Title) Explainable Deep Neural Network for Anesthetic Treatment Recommendation
ÀúÀÚ(Author) ÀÌÀç±Ô   À̻󿱠  Jaekyu Lee   Sangyub Lee   ¼º¼öÁø   ±Ç¼ö¹ü   À±Áö¿í   ¿ÀÁø¿µ   Â÷Á¤¿ø   Su-Jin Seong   Soo-Bum Kwon   Ji-Uk Yoon   Jin-Yong Oh   Jeong-Won Cha  
¿ø¹®¼ö·Ïó(Citation) VOL 26 NO. 12 PP. 0550 ~ 0555 (2020. 12)
Çѱ۳»¿ë
(Korean Abstract)
º» ³í¹®¿¡¼­´Â ¼ö¼ú Áß È¯ÀÚ¿¡ ´ëÇÑ 17°³ÀÇ »ýüÁ¤º¸(Vital sing)¿¡ ´ëÀÀÇÏ¿© ÁÖ¾îÁø óġ Èĺ¸ Áß ÃÖÀûÀÇ Ã³Ä¡ ¹æ¹ýÀ» ÃßõÇÏ´Â ¸¶Ãë¿ë óġ Ãßõ ¸ðµ¨ ±¸Ãà ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. ÀÇÇÐ ºÐ¾ß ½Ã½ºÅÛÀÇ °á°ú´Â ÃÖÁ¾ÀûÀ¸·Î »ç¶÷¿¡°Ô Àû¿ëµÇ±â ¶§¹®¿¡ ¸ðµ¨ÀÇ °á°ú¿¡ ´ëÇÑ ½Å·Ú¼º È®º¸°¡ Áß¿äÇÏ´Ù. Á¦¾È ¸ðµ¨Àº °¡ÁßÄ¡ °ö¿¡¼­ ÀÌ·ç¾îÁö´Â ÇÕ ¿¬»êÀ» max-poolingÀ¸·Î ´ëüÇØ °¢ ÀÔ·Â ÀÚÁú¿¡ ´ëÇÑ °¡ÁßÄ¡¸¦ µ¶¸³ÀûÀ¸·Î °è»êÇÑ´Ù. µû¶ó¼­ ¸ðµ¨Àº gradient descent¸¦ ÀÌ¿ëÇÏ¿© ÇнÀµÉ ¼ö ÀÖÀ¸¸ç, ÃßõµÈ °á°ú¿¡ ´ëÇÑ ±Ù°Å¸¦ Á¦½ÃÇÒ ¼ö ÀÖ´Ù. ¾ç»êºÎ»ê´ëº´¿øÀÇ µ¥ÀÌÅ͸¦ ÀÌ¿ëÇÑ ¿¹ºñ ½ÇÇèÀ» ÅëÇØ °¡´É¼ºÀ» º¸À̸ç, Ãß°¡·Î ¼öÁýÇÑ Á¤Çü µ¥ÀÌÅÍ¿¡ Àû¿ëÇÏ¿© È¿¿ë¼ºÀ» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
In this paper, we propose a method for constructing an anesthesia treatment recommendation model that suggests an optimal treatment among given treatment candidates in response to 17 vital information regarding a patient during surgery. Since the results of the model in the medical field are finally applied to humans, it is crucial to ensure the reliability of the model results. The proposed model replaces the summation with max-pooling to independently calculate the weights for each input feature. Thus, the model can be trained using gradient descent and can provide a basis for the recommended results. We show the possibility through preliminary experiments using data from Pusan National University Yangsan Hospital, and apply the method to the additional collected structured data to confirm its effectiveness.
Å°¿öµå(Keyword) Â÷·® Á¶¸íÁ¦¾î   ÀüÀÚ±âÀå   ¼Õµ¿ÀÛ ÀνĠ  Á¶¸í Á¦¾î   Â÷·® »ç¿ëÀÚ ÀÎÅÍÆäÀ̽º   vehicle lighting control   electromagnetic field   hand gesture recognition   lighting control   in-vehicle user interface   µö·¯´×   ¼³¸í °¡´ÉÇÑ ÀΰøÁö´É   ¸¶Ãë óġ Ãßõ   Á¤Çü µ¥ÀÌÅÍ   deep learning   explainable AI   anesthetic treatment recommendation   structured data  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå