• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2021년도 춘계학술대회
• /
• pp.134-136
• /
• 2021

인공지능 알고리즘을 이용한 유방암의 조기진단에 관련된 연구는 최근들어 활발하게 진행되고 있다. 이는 연구용으로 공개된 초음파 유방 이미지를 활용하여 다양하게 개발되고 있으나, 사용자의 목적에 맞는 처리 속도 및 정확도 등에 다양한 한계점을 보인다. 이러한 문제를 해결하기 위해, 본 논문에서는 ImageNet에서 학습된 ResNet 모델을 현미경 기반 암세포 이미지에서 활용이 가능한 다단계 전이 학습을 제안하고, 이를 다시 전이 학습하여 초음파 유방암 영상을 양성 및 악성으로 분류하는 실험을 진행하였다. 실험을 위한 영상은 양성과 악성이 포함된 250장의 유방암 초음파 영상과 27,200장의 암 세포주 영상으로 구성되었다. 제안된 다단계 전이 학습 알고리즘은 초음파 유방암 영상을 분류하였을 때 96% 이상의 정확도를 보였으며, 향후 암 세포주 및 실시간 영상처리 등의 추가를 통해 보다 높은 활용도와 정확도를 보일 것으로 기대한다.

• 여성건강간호학회지
• /
• 제1권1호
• /
• pp.138-152
• /
• 1995

The condition of nutrition during infancy will greatly affect infants' physical, emotional growth, especially breast feeding is important in their growth and development, and emotional stability, too. Despite such advantages of breast-feeding, its rate has continued to fall year after year in Korea : the 95% rate in 1960's has fallen to 25.4% in 1990. It is known that such a downfall of breast-feeding rate is associated with various factors. The purpose of the study is to examine mother's and nurses' in nursery perception of the impediment factors of breast-feeding : to compare those between of breast-feeding : to compare those between two groups : to provide fundamental data for developing strategies for increasing breast-feeding. The subjects were 45 new mothers from one hospital and the same number of nurses sampled from 3 university hospitals, in Seoul. The data were collected for 11 days from April 12 to April 23, 1994 and a questionnaire was developed based on the interview with 14 nurses and 10 mothers and the literature reviews. Liker's Five-Point Scale was used as measurement. The Data were analyzed using SPSS / PC and descriptive statistics, t-test, ANOVA. The results of study are as follows : 1) Most new mothers get the information about breast-feeding from their mothers, books or newspaper, and relatives in rank than professionals(nurses or doctors). 2) The impediment factors the breast-feeding are categorized as mothers, hospital system, medical personnels and social factors. The most frequently mentioned impediment factors from mothers are 'Difficulty by operational wounds'(3.13), 'Lack of will for breast-feeding'(3.47), 'Insufficient rest and sleep'(3.52) and 'Opposition from husband or his family members'(4.77)in rank. On the other hand those factors from nurses are 'Inadequate nipple condition'(2.37), 'Decreasing milk secretion given medicine after operation'(2.63), 'Mothers knowledge deficit by poorly prepared education'(2.79) and 'Mothers abhorance of breast-feeing'(3.87) in rank. 3) As for the hospital system, the highly perceived impediment factors by mothers are 'Seperation of baby from mother after birth'(2.78), 'Lack of space for breast-feeding in the hospital'(2.93), 'Lack of facility for informing the time for breast-feeding'(3.18) and 'Because of babys' preference artificial nipple by being accustomed to it in hospital(3.97), in rank. Meanwhile, those from nurses are "Seperation of baby from mother after birth",(1.92), "Inconsistency between hospital's nursing time and mother's breast secretion time" scretion time(2.97), "Lack of space for breast-feeding in the hospital"(3.39), and "Lack of facility for informing the time for breast-feeding"(3.74) 4) As for the medical personnels, the highestly perceived Impediment factor from mothers in "Lack of professional nurses for breast-feeding"(2.96), and the lowestly perceived one is "Doctors' reluctance"(4.75). Nurses perceived same as mothers, too. 5) As for the social factors, the highestly perceived impediment factor by mothers and nurses is "Inconvience of social activities"(2.83) and the lowestly ones are "The sense of self-sacrifice"(4.22) by mothers, and "The sense of old fashioned"(4.13) by nurses. 6) The difference of the perception of impediment factors between mothers and nurses is statistically significant only in mother factor.

• 여성건강간호학회지
• /
• 제29권3호
• /
• pp.160-178
• /
• 2023

Purpose: Since technology-based interventions can facilitate convenient access to healthcare for women with breast cancer, it is crucial to understand innovative approaches to maintaining the effectiveness of these interventions. Therefore, we conducted a systematic review of technology-based self-management interventions for women with breast cancer in six countries. We analyzed the characteristics of these interventions and examined their diverse health outcomes. Methods: Six databases were systematically searched to extract research articles using the keywords "breast cancer," "technology," and "self-management." The search was carried out up until June 12, 2023. From the 1,288 studies retrieved from the database search, 10 eligible papers were identified based on inclusion/exclusion criteria. Two authors independently extracted and compared the data from these articles, resolving any discrepancies through discussion. Results: Most of the 10 studies utilized web- or mobile-based technology, and one used artificial intelligence-based technology. Among the 12 health-related outcome variables, quality of life and symptom distress were the most frequently mentioned, appearing in six articles. Furthermore, an analysis of the intervention programs revealed a variety of common constructs and the involvement of managers in the self-management intervention. Conclusion: Incorporating key components such as self-management planning, diary keeping, and communication support in technology-based interventions could significantly improve the self-management process for breast cancer survivors. The practical application of technology has the potential to empower women diagnosed with breast cancer and improve their overall quality of life, by providing timely and sustainable interventions, and by leveraging available resources and tools.

• Korean Journal of Radiology
• /
• 제23권5호
• /
• pp.505-516
• /
• 2022

Objective: To evaluate whether artificial intelligence (AI) for detecting breast cancer on mammography can improve the performance and time efficiency of radiologists reading mammograms. Materials and Methods: A commercial deep learning-based software for mammography was validated using external data collected from 200 patients, 100 each with and without breast cancer (40 with benign lesions and 60 without lesions) from one hospital. Ten readers, including five breast specialist radiologists (BSRs) and five general radiologists (GRs), assessed all mammography images using a seven-point scale to rate the likelihood of malignancy in two sessions, with and without the aid of the AI-based software, and the reading time was automatically recorded using a web-based reporting system. Two reading sessions were conducted with a two-month washout period in between. Differences in the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, and reading time between reading with and without AI were analyzed, accounting for data clustering by readers when indicated. Results: The AUROC of the AI alone, BSR (average across five readers), and GR (average across five readers) groups was 0.915 (95% confidence interval, 0.876-0.954), 0.813 (0.756-0.870), and 0.684 (0.616-0.752), respectively. With AI assistance, the AUROC significantly increased to 0.884 (0.840-0.928) and 0.833 (0.779-0.887) in the BSR and GR groups, respectively (p = 0.007 and p < 0.001, respectively). Sensitivity was improved by AI assistance in both groups (74.6% vs. 88.6% in BSR, p < 0.001; 52.1% vs. 79.4% in GR, p < 0.001), but the specificity did not differ significantly (66.6% vs. 66.4% in BSR, p = 0.238; 70.8% vs. 70.0% in GR, p = 0.689). The average reading time pooled across readers was significantly decreased by AI assistance for BSRs (82.73 vs. 73.04 seconds, p < 0.001) but increased in GRs (35.44 vs. 42.52 seconds, p < 0.001). Conclusion: AI-based software improved the performance of radiologists regardless of their experience and affected the reading time.

• Archives of Plastic Surgery
• /
• 제35권3호
• /
• pp.279-282
• /
• 2008

Purpose: It is known that the chronic absence of unilateral breast can cause spine curvature. The artificial breast manufactures take up the position of that possibility. This study was designed to evaluate the influence of the mastectomy on the spine and appearance of scoliosis among women who wanted delayed breast reconstruction. ted delayed breast reconstruction. Methods: The study population consisted of 47 women who underwent delayed breast reconstruction at our Department of Plastic Surgery from April 2001 to May 2007. The whole spine anteroposterior and lateral X-ray was taken to evaluate the Cobb's angle. As a general rule a Cobb angle of 10 is regarded as a minimum angulation to define scoliosis. We evaluated Cobb's angle and drew a correlation between the duration of the mastectomy state and the weight of the mastectomy specimen. Results: There were no family history of scoliosis, and no numbness or weakness in the upper or lower extremities. They had normal reflex and experienced no tenderness of the perispinal area. The Spearman Correlation Coefficient between Cobb's angle and the period that took time from the mastectomy to the X-rays and analysis between Cobb's angle and specimen weight was 0.032 and-0.115. there were no correlation between Cobb's angle and time, or between Cobb's angle and specimen weight. Conclusion: lthough the patients attribute their back pain and distorted posture to having received an one-sided mastectomy, it is unlikely that one-sided mastectomy causes spinal deformity and scoliosis.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2007년도 한국지능정보시스템학회
• /
• pp.389-398
• /
• 2007

Case-based reasoning (CBR) is one of the most popular prediction techniques for medical diagnosis because it is easy to apply, has no possibility of overfitting, and provides a good explanation for the output. However, it has a critical limitation - its prediction performance is generally lower than other artificial intelligence techniques like artificial neural networks (ANNs). In order to obtain accurate results from CBR, effective retrieval and matching of useful prior cases for the problem is essential, but it is still a controversial issue to design a good matching and retrieval mechanism for CBR systems. In this study, we propose a novel approach to enhance the prediction performance of CBR. Our suggestion is the simultaneous optimization of feature weights, instance selection, and the number of neighbors that combine using genetic algorithms (GAs). Our model improves the prediction performance in three ways - (1) measuring similarity between cases more accurately by considering relative importance of each feature, (2) eliminating redundant or erroneous reference cases, and (3) combining several similar cases represent significant patterns. To validate the usefulness of our model, this study applied it to a real-world case for evaluating cytological features derived directly from a digital scan of breast fine needle aspirate (FNA) slides. Experimental results showed that the prediction accuracy of conventional CBR may be improved significantly by using our model. We also found that our proposed model outperformed all the other optimized models for CBR using GA.

• 대한간호학회지
• /
• 제7권1호
• /
• pp.30-38
• /
• 1977

Breast cancer is one of the most feared health problems in women ; Recent studies revealed that it had come up to be the second most in this country and high prevalent disease in the western countries among breast disease in women. However, early detection of the cancer mass is known to be easier than in many other malignancies. This study was performed to investigate the various characteristics of patients of breast cancer ; by the structural variables, menstrual, marital, and child bearing, and also their understandings and attitude towards the disease. A hundred any in- and out- patients of 51. Mary's Hospital and National Atomic Institute, Seoul were sampled. Data were gathered through direct interview by the researcher from February 1976 to January 1977 and the clinical records were used as references. Results are as follows ; 1. Breast cancer revealed to be most prevalent in the forties ; average age of 45.2 years. 2. The average age of menarche revealed to be 15.4 years ; the largest group were the 16-20 years (N=75, 68.2%). In 55 cases (50%) menstruation were normal, 38 (34.5%) postmenopausal and only in 17 (15.5%), menstruation revealed irregularity. 3. The average marital age revealed to be 22.3 years ; the largest group were 21-25 group (N=43, 39.1%). The average duration of marital life revealed to be 24.7 years ; 11-20 years group were the largest (M=34, 30.9%). 4. Most of the patients revealed to have pregnancy experiences(N=100, 90.9%) ; the average rate of experience were 5, 3 times the largest group were 3.4 times group(N= 32, 29.1%). 54 patients (49.1%) revealed to have had abortion experience ; the average were 3.4 times. 5. The largest group(N=77, 70%) had been breast feeding : followed by mixed feeding (N=12, 10.9%) and artificial feeding(N=10, 9.1%). 6. Personal health history revealed that in 20 patients (18.2%) revealed to have the past history of Purulent mastitis, 5 patients(4.5%) of breast cancer and 3 patients(2.7%) of uttering cancer family history. 7. In the one half (N=56, 50.9%, they had some information about breast cancer :27 (24.5%) by mass media, 12 (10.9%) through personal contacts and 17 (15.5%) were not able to classify the source of information. 8. In 55 cases (50%) the canoe, mass were discovered incidentally, in 39 cases (35.5%) by manual detection by self, in 10 cases (9%) by others and in 6 cases(5.5%) by observing subjective symptoms. 9. The average duration lapsed between the discovery of cancer mass and the visit to the hospital revealed to be 9.4 month. Chief reason for the delay revealed to be the non-chaplaincy due to the absence of pain(N=50, 45.5 %) followed by the administration of herb and commercial medication (N=19, 17.3%). 10. The left side breast was more affected than the right side breast, represent by 60 cases in the left and 39 cases in the right. The most frequent site of the breast cancer was the upper- outer quadrant in 53 cases (47.7%), and followed by the center in 20 cases(18 %), and the upper inner quadrant, in 19 cases (17.1%). There was / cases of bilateral carcinoma. The most prominent symptom was painless mass.

• 한국컴퓨터정보학회논문지
• /
• 제28권11호
• /
• pp.1-11
• /
• 2023

유방암은 전 세계적으로 여성들 대다수에게 가장 두려워하는 질환이다. 오늘날 데이터의 증가와 컴퓨팅 기술의 향상으로 머신러닝(machine learning)의 효율성이 증대되어 암 검출 및 진단 등에 중요한 역할을 하고 있다. 딥러닝(deep learning)은 인공신경망(artificial neural network, ANN)을 기반으로 하는 머신러닝 기술의 한 분야로 최근 여러 분야에서 성능이 급속도로 개선되어 활용 범위가 확대되고 있다. 본 연구에서는 유방암 분류를 위해 전이학습(transfer learning) 기반 DNN(Deep Neural Network)과 SVM(support vector machine)의 구조를 결합한 DNN-SVM Hybrid 모형을 제안한다. 전이학습 기반 제안된 모형은 적은 학습 데이터에도 효과적이고, 학습 속도도 빠르며, 단일모형, 즉 DNN과 SVM이 가지는 장점을 모두 활용 가능토록 결합함으로써 모형 성능이 개선되었다. 제안된 DNN-SVM Hybrid 모형의 성능평가를 위해 UCI 머신러닝 저장소에서 제공하는 WOBC와 WDBC 유방암 자료를 가지고 성능실험 결과, 제안된 모형은 여러 가지 성능 척도 면에서 단일모형인 로지스틱회귀 모형, DNN, SVM 그리고 앙상블 모형인 랜덤 포레스트보다 우수함을 보였다.

• 한국융합학회논문지
• /
• 제10권2호
• /
• pp.57-66
• /
• 2019

본 연구의 목적은 유방질환 발생여부에 따른 생활습관과 천연 에스트로겐이 함유된 건강기능 식품 섭취실태 차이를 확인하여 유방질환 발생 위험요인을 파악하기 위함이다. 본 연구의 대상자는 20세 이상 60세사이의 J시 소재 M여성 전문병원에 건강검진을 위한 내원한 여성 중 유방질환이 있는 군을 질환군 85명, 대조군 93명 등 총 178명을 대상으로 하였다. 본 연구결과 인공 유산경험이 있는 경우가 유방질환 진단을 받을 가능성이 3.04배, 백미 섭취군이 유방질환 진단을 받을 가능성이 4.76배, 스트레스 해소방법의 취미생활은 하지 않는 경우가 유방질환 질환 군으로 분류될 가능성이 2.38배 증가하였다. 본 연구결과 인공 유산경험이 있고, 백미를 즐겨 먹으며, 취미생활을 하지 않는 대상자는 유방질환을 진단 받을 가능성이 증가하였다. 이와 같은 특성을 반영한 일반인을 위한 유방질환 예방홍보와 교육 프로그램이 필요할 것이다.

• Child Health Nursing Research
• /
• 제12권3호
• /
• pp.386-397
• /
• 2006

Purpose: This study was performed to compare the difference of maternal attachment and the maternal role confidence between mother who feeds the child with mother's milk in sanitary pack by a nurse instead of her and mother who feeds the child with artificial milk. Methods: The subjects of this study consisted of mothers of premature baby who was in NICU and can not be fed with mother's milk directly. In the sample, 21 mothers were the breast feeding group and 20 were the bottle feeding group. Data were collected from April 3, 2004 to November 2, 2005, and were analyzed using SPSS WIN 10.0. Results: The degree of maternal attachment and maternal role confidence of the breast milk feeding group was higher than that of the bottle feeding group. In accordance with general characteristics, the difference was found in maternal attachment and maternal role confidence both breast feeding group and bottle feeding group. Conclusion: More systematic nursing mediation is required for the lactation of mother's milk in sanitary pack is planned to do positive interaction between mother and the child, which has an influence on the formation of maternal attachment and the of maternal role confidence after hospitalization.