Deep learning algorithm for predicting preterm birth in the case of threatened preterm labor admissions using transvaginal ultrasound.

PURPOSE: Preterm birth presents a major challenge in perinatal care, and predicting preterm birth remains a major challenge. If preterm birth cases can be accurately predicted during pregnancy, preventive interventions and more intensive prenatal monitoring may be possible. Deep learning has the capability to extract image parameters or features related to diseases. We constructed a deep learning model to predict preterm births using transvaginal ultrasound images. METHODS: Patients who were hospitalized for threatened preterm labor or shortened cervical length were enrolled. We used images of the cervix obtained via transvaginal ultrasound examination at admission to predict cases of preterm birth. We used convolutional neural networks (CNNs) and Vision Transformer (Vit) for the model construction. We compared the prediction performance of deep learning models with two human experts. RESULTS: A total of 59 patients were enrolled in the study, including 30 cases in the preterm group and 29 cases in the full-term group. Statistical analysis of clinical variables including cervical length showed no significant differences between the two groups. For accuracy, the best CNN model had the highest accuracy of 0.718 with an area under the curve (AUC) of 0.704, followed by Vision Transformer with accuracy of 0.645 and AUC of 0.587. The accuracy of two human experts was 0.465 and 0.517, respectively. CONCLUSIONS: Deep learning models have important implications for extraction of features that provide more accurate assessment of preterm birth than traditional visual assessment by the human eye.

A multimodal deep learning model for predicting severe hemorrhage in placenta previa.

Placenta previa causes life-threatening bleeding and accurate prediction of severe hemorrhage leads to risk stratification and optimum allocation of interventions. We aimed to use a multimodal deep learning model to predict severe hemorrhage. Using MRI T2-weighted image of the placenta and tabular data consisting of patient demographics and preoperative blood examination data, a multimodal deep learning model was constructed to predict cases of intraoperative blood loss > 2000 ml. We evaluated the prediction performance of the model by comparing it with that of two machine learning methods using only tabular data and MRI images, as well as with that of two human expert obstetricians. Among the enrolled 48 patients, 26 (54.2%) lost > 2000 ml of blood and 22 (45.8%) lost < 2000 ml of blood. Multimodal deep learning model showed the best accuracy of 0.68 and AUC of 0.74, whereas the machine learning model using tabular data and MRI images had a class accuracy of 0.61 and 0.53, respectively. The human experts had median accuracies of 0.61. Multimodal deep learning models could integrate the two types of information and predict severe hemorrhage cases. The model might assist human expert in the prediction of intraoperative hemorrhage in the case of placenta previa.

Preliminary Results of Deep Learning Approach for Preoperative Diagnosis of Ovarian Cancer Based on Pelvic MRI Scans.

BACKGROUND/AIM: To predict the pathological diagnosis of ovarian tumors using preoperative MRI images, using deep learning models. PATIENTS AND METHODS: A total of 185 patients were enrolled, including 40 with ovarian cancers, 25 with borderline malignant tumors, and 120 with benign tumors. Using sagittal and horizontal T2-weighted images (T2WI), we constructed the pre-trained convolutional neural networks to predict pathological diagnoses. The performance of the model was assessed by precision, recall, and F1-score on macro-average with 95% confidence interval (95%CI). The accuracy and area under the curve (AUC) were also assessed after binary transformation by the division into benign and non-benign groups. RESULTS: The macro-average accuracy in the three-class classification was 0.523 (95%CI=0.504-0.544) for sagittal images and 0.426 (95%CI=0.404-0.446) for horizontal images. The model achieved a precision of 0.63 (95%CI=0.61-0.66), recall of 0.75 (95%CI=0.72-0.78), and F1 score of 0.69 (95%CI=0.67-0.71) for benign tumor. Regarding the discrimination between benign and non-benign tumors, the accuracy in the binary-class classification was 0.628 (95%CI=0.592-0.662) for sagittal images and AUC was 0.529 (95%CI=0.500-0.557). CONCLUSION: Using deep learning, we could perform pathological diagnosis from preoperative MRI images.

Prediction of preterm birth using artificial intelligence: a systematic review.

Preterm birth is the leading cause of neonatal death. It is challenging to predict preterm birth. We elucidated the state of artificial intelligence research on the prediction of preterm birth, clarifying the predictive values and accuracy. We performed a systematic review using three databases (PubMed, Web of Science, and Scopus) in August 2020, with keywords as 'artificial intelligence,' 'deep learning,' 'machine learning,' and 'neural network' combined with 'preterm birth'. We included 22 publications between 2010 and 2020. Regarding the predictive values, electrohysterogram images were mostly used, followed by the biological profiles, the metabolic panel in amniotic fluid or maternal blood, and the cervical images on the ultrasound examination. The size of dataset in most studies was hundred cases and too small for learning, although only three studies used the medical database over a hundred thousand cases. The accuracy was better in the studies using the metabolic panel and electrohysterogram images. Impact statementWhat is already known on this subject? Preterm birth is the leading cause of newborn morbidity and mortality. Presently, the prediction of preterm birth in individual cases is still challenging.What the results of this study add? Using artificial intelligence such as deep learning and machine learning models, clinical data could lead to accurate prediction of preterm birth.What the implications are of these findings for clinical practice and/or further research? The size of the datasets was too small for the models using artificial intelligence in the previous studies. Big data should be prepared for the future studies.

Machine learning approach for the prediction of postpartum hemorrhage in vaginal birth.

Postpartum hemorrhage is the leading cause of maternal morbidity. Clinical prediction of postpartum hemorrhage remains challenging, particularly in the case of a vaginal birth. We studied machine learning models to predict postpartum hemorrhage. Women who underwent vaginal birth at the Tokyo Women Medical University East Center between 1995 and 2020 were included. We used 11 clinical variables to predict a postpartum hemorrhage defined as a blood loss of > 1000 mL. We constructed five machine learning models and a deep learning model consisting of neural networks with two layers after applying the ensemble learning of five machine learning classifiers, namely, logistic regression, a support vector machine, random forest, boosting trees, and decision tree. For an evaluation of the performance, we applied the area under the curve of the receiver operating characteristic (AUC), the accuracy, false positive rate (FPR) and false negative rate (FNR). The importance of each variable was evaluated through a comparison of the feature importance calculated using a Boosted tree. A total of 9,894 patients who underwent vaginal birth were enrolled in the study, including 188 cases (1.9%) with blood loss of > 1000 mL. The best learning model predicted postpartum hemorrhage with an AUC of 0.708, an accuracy of 0.686, FPR of 0.312, and FNR of 0.398. The analysis of the importance of the variables showed that pregnant gestation of labor, the maternal weight upon admission of labor, and the maternal weight before pregnancy were considered to be weighted factors. Machine learning model can predict postpartum hemorrhage during vaginal delivery. Further research should be conducted to analyze appropriate variables and prepare big data, such as hundreds of thousands of cases.

Artificial intelligence in gynecologic cancers: Current status and future challenges - A systematic review.

OBJECTIVE: Over the past years, the application of artificial intelligence (AI) in medicine has increased rapidly, especially in diagnostics, and in the near future, the role of AI in medicine will become progressively more important. In this study, we elucidated the state of AI research on gynecologic cancers. METHODS: A search was conducted in three databases-PubMed, Web of Science, and Scopus-for research papers dated between January 2010 and December 2020. As keywords, we used "artificial intelligence," "deep learning," "machine learning," and "neural network," combined with "cervical cancer," "endometrial cancer," "uterine cancer," and "ovarian cancer." We excluded genomic and molecular research, as well as automated pap-smear diagnoses and digital colposcopy. RESULTS: Of 1632 articles, 71 were eligible, including 34 on cervical cancer, 13 on endometrial cancer, three on uterine sarcoma, and 21 on ovarian cancer. A total of 35 studies (49%) used imaging data and 36 studies (51%) used value-based data as the input data. Magnetic resonance imaging (MRI), computed tomography (CT), ultrasound, cytology, and hysteroscopy data were used as imaging data, and the patients' backgrounds, blood examinations, tumor markers, and indices in pathological examination were used as value-based data. The targets of prediction were definitive diagnosis and prognostic outcome, including overall survival and lymph node metastasis. The size of the dataset was relatively small because 64 studies (90%) included less than 1000 cases, and the median size was 214 cases. The models were evaluated by accuracy scores, area under the receiver operating curve (AUC), and sensitivity/specificity. Owing to the heterogeneity, a quantitative synthesis was not appropriate in this review. CONCLUSIONS: In gynecologic oncology, more studies have been conducted on cervical cancer than on ovarian and endometrial cancers. Prognoses were mainly used in the study of cervical cancer, whereas diagnoses were primarily used for studying ovarian cancer. The proficiency of the study design for endometrial cancer and uterine sarcoma was unclear because of the small number of studies conducted. The small size of the dataset and the lack of a dataset for external validation were indicated as the challenges of the studies.

Learning Curve of Robotic-assisted Hysterectomy With Pelvic Lymphadenectomy for Early Stage Endometrial Cancer: Analysis of 81 Cases.

BACKGROUND/AIM: The purpose of this study was to evaluate the learning curve of robotic hysterectomy and pelvic lymphadenectomy for early-stage endometrial carcinoma. PATIENTS AND METHODS: A retrospective chart review was performed on the first 81 surgeries performed by a single surgeon. The 81 cases were divided into three groups; 4 subgroups of 20 cases each, 3 subgroups of 27 cases each, and 2 subgroups of 40 cases each. The surgical outcomes in each group were analyzed, using operative time, estimated blood loss, and the number of lymph nodes resected. RESULTS: The median operating time, estimated blood loss, and number of pelvic lymph nodes were 147 min, 50 g and 23, respectively. The estimated blood loss improved over time significantly, when dividing by every 27 and 40 cases. No statistical significance was shown regarding operative time and the number of lymph nodes. CONCLUSION: Approximately, 30 cases were needed to gain proficiency in the surgical technique.

The application of machine learning for predicting recurrence in patients with early-stage endometrial cancer: a pilot study.

OBJECTIVE: Most women with early stage endometrial cancer have a favorable prognosis. However, there is a subset of patients who develop recurrence. In addition to the pathological stage, clinical and therapeutic factors affect the probability of recurrence. Machine learning is a subtype of artificial intelligence that is considered effective for predictive tasks. We tried to predict recurrence in early stage endometrial cancer using machine learning methods based on clinical data. METHODS: We enrolled 75 patients with early stage endometrial cancer (International Federation of Gynecology and Obstetrics stage I or II) who had received surgical treatment at our institute. A total of 5 machine learning classifiers were used, including support vector machine (SVM), random forest (RF), decision tree (DT), logistic regression (LR), and boosted tree, to predict the recurrence based on 16 parameters (age, body mass index, gravity/parity, hypertension/diabetic, stage, histological type, grade, surgical content and adjuvant chemotherapy). We analyzed the classification accuracy and the area under the curve (AUC). RESULTS: The highest accuracy was 0.82 for SVM, followed by 0.77 for RF, 0.74 for LR, 0.66 for DT, and 0.66 for boosted trees. The highest AUC was 0.53 for LR, followed by 0.52 for boosted trees, 0.48 for DT, and 0.47 for RF. Therefore, the best predictive model for this analysis was LR. CONCLUSION: The performance of the machine learning classifiers was not optimal owing to the small size of the dataset. The use of a machine learning model made it possible to predict recurrence in early stage endometrial cancer.

Artificial Intelligence in Ovarian Cancer Diagnosis.

BACKGROUND/AIM: This study aimed to use artificial intelligence (AI) to predict the pathological diagnosis of ovarian tumors using patient information and data from preoperative examinations. PATIENTS AND METHODS: A total of 202 patients with ovarian tumors were enrolled, including 53 with ovarian cancer, 23 with borderline malignant tumors, and 126 with benign ovarian tumors. Using 5 machine learning classifiers, including support vector machine, random forest, naive Bayes, logistic regression, and XGBoost, we derived diagnostic results from 16 features, commonly available from blood tests, patient background, and imaging tests. We also analyzed the importance of 16 features on the prediction of disease. RESULTS: The highest accuracy was 0.80 in the machine learning algorithm of XGBoost. The evaluation of importance of the features showed different results among the correlation coefficient of the features, the regression coefficient, and the features importance of random forest. CONCLUSION: AI could play a role in the prediction of pathological diagnosis of ovarian cancer from preoperative examinations.

Early Recurrence of Ovarian Cancer during Pregnancy after Primary Staging Surgery in the First Trimester.

Cancer during pregnancy is rare. However, even during pregnancy, there is the possibility of recurrence for the patients. We present the case of recurrence of ovarian cancer during pregnancy regardless of primary staging surgery performed in the first trimester of the same pregnancy. The patient was a 29-year-old woman who underwent fertility-sparing surgery at 15 weeks of pregnancy for ovarian cancer (mucinous adenocarcinoma, FIGO stage IC). Omitting adjuvant chemotherapy during pregnancy, we continued the prenatal checkups in the outpatient. At 31 weeks of gestation, massive ascites emerged and oliguria/anuria developed acutely. We performed emergent cesarean section, diagnosing acute kidney injury during pregnancy. On surgical finding, there were a number of 1 cm sized nodules in the small bowel wall and peritoneum. The infant was appropriate for gestational age without any abnormalities. Oliguria continued due to rapid accumulation of ascites in the early postpartum period. After two cycles of chemotherapy, ascites decreased gradually and the markers gradually decreased. However, after six courses of chemotherapy, she suddenly complained of nausea and anorexia. CT imaging showed cancerous ileus and ascites fluids. The patient chose palliative care. Even in the case of nonadvanced cancer, it has the potential to be an extremely aggressive malignancy under the irregular hormonal environment of pregnancy.

Impact of uterine weight on robotic hysterectomy: Analysis of 500 cases in a single institute.

BACKGROUNDS: Robotic surgeries have been used frequently for benign diseases in gynecology. However, the advantage of robotic surgery for huge uterus is unclear. METHODS: We analyzed surgical outcomes of 527 patients who underwent robotic hysterectomies for benign diseases, separating uterine sizes into five groups by every 250 g. RESULTS: Median operative time in the five groups was 123 minutes (<250 g), 130 minutes (250-500 g), 144 minutes (500-750 g), 180 minutes (750-1000 g), and 170 minutes (>1000 g). Median estimated blood loss was 50, 100, 100, 200, and 400 mL in the five groups, respectively. The incidence of intraoperative complications did not correlate with uterine weight. CONCLUSIONS: Operative time, estimated blood loss, and the incidence of conversion to laparotomy increased with uterine size during robotic hysterectomies, especially evident in a uterus >750 g.

Allergy-like reactions to methylene blue following laparoscopic chromopertubation: A systematic review of the literature.

Methylene blue is commonly used as a tracer in sentinel lymph node mapping for many malignant diseases or chromopertubation during gynecologic laparoscopy. In contrast with other blue dyes such as patent blue V or isosulfan blue, methylene blue rarely causes an allergy-like reaction in patients undergoing sentinel lymph node mapping. However, in chromopertubation, some cases of allergy-like reaction to methylene blue have been reported; these comprise two types: an allergic reaction and methemoglobinemia. In this study, a systematic literature review of allergy-like reactions caused by methylene blue dye following laparoscopic chromopertubation was conducted. A search was conducted in PUBMED, Web of Science, and Scopus from inception until June 2018, using the terms: "methylene blue", "complication", "allergic", "hypersensitive", "lung/pulmonary edema"," methemoglobinemia", "anaphylactic shock", "chromopertubation", "pertubation", "laparoscopic", and "laparoscopy". Ultimately, the eligibility criteria were fulfilled by only 12 case reports. Among 13 cases including our case of severe anaphylactic shock after chromopertubation, allergic reactions were diagnosed in four cases, methemoglobinemia in six, and there was no confirmed diagnosis in three cases; the clinical course consisted of skin changes, blue discoloration of body fluids, respiratory failure, and hemodynamic failure, regardless of the underlying diagnoses. Regarding diagnosis, methemoglobinemia was confirmed with co-oximetry (spectrophotometry). First-line therapy included supportive care for both cases of allergic reactions and methemoglobinemia.

Comparison of Electrosurgical Devices for Cervical Conization: Novel Monopolar Scalpel (VIO) Versus Ultrasonic Scalpel.

OBJECTIVE: The aim of this study was to compare the outcomes associated with the use of a novel monopolar scalpel with those associated with the use of an ultrasonic scalpel for cervical conization of cervical intraepithelial neoplasia. MATERIALS AND METHODS: We conducted a retrospective cross-sectional study in patients treated in our institute between April 2010 and March 2017. We used either the VIO monopolar scalpel (VIO) or Harmonic ultrasonic scalpel (HS) for cervical conization. We analyzed operative outcomes, postoperative complications, and pathological findings associated with the use of the 2 devices. RESULTS: In 500 patients treated with cervical conization, VIO and HS were used in 249 and 251 patients, respectively. No significant difference in patient background was found between the groups. The mean operative time was shorter with VIO than with HS (18.2 min vs. 27.4 min). The mean estimated blood loss was greater with VIO (7.2 g vs. 3.1 g), but the postoperative bleeding rate was higher with HS (5% vs. 20%). Regarding other complications, cervical stenosis was only noted with VIO (4 cases, 1.6%). The positive margin (11% vs. 16%) and positive endocervical curettage rates (7% vs 10%) were not significantly different between the groups. No significant differences were also found in the pathological results and need for additional treatment (the rate of the additional treatment: 20% vs. 23%). CONCLUSIONS: Considering short operating time and less postoperative bleeding, VIO was preferred to HS. However, the excessive coagulation in VIO is considered to lead to cervical stenosis.

Malignant Transformation of Mature Cystic Teratoma: Is Squamous Cell Carcinoma Different From the Other Types of Neoplasm?

OBJECTIVES: Malignant transformation of mature cystic teratoma (MCT) is rare. Unlike squamous cell carcinoma (SCC) in MCT, the other types of neoplasm in MCT have not been discussed in publications. We analyzed the clinical characteristics and prognosis of the other types of neoplasm (non-SCC) compared with those of SCC. METHODS: A systematic literature search of literature published from 2000 to 2017 was conducted in PubMed, Web of Science, and Scopus. We reviewed case series that included all pathological types of malignant transformation. RESULTS: A total of 155 cases from 15 case series, including our cases, were included. Of the cases, 90 (58%) were SCC and 65 (42%) were non-SCC, including adenocarcinoma, carcinoid tumor, thyroid carcinoma, sarcoma, adenosquamous carcinoma, melanoma, sebaceous carcinoma, oligodendroglioma, signet ring cell carcinoma, and transitional cell carcinoma, in descending order of frequency. The mean ages of patients with SCC and non-SCC were 50.5 and 48.9 years, respectively. The mean tumor sizes were 14.7 cm in SCC and 13.9 cm in non-SCC. Surgical approaches were similar. First-line chemotherapy for epithelial ovarian cancers was the most commonly used regimen in SCC and non-SCC. Overall survival did not differ significantly, showing better prognosis in stage I and poor prognosis in stages II, III, and IV. A difference in overall survival was observed among pathological types of non-SCC. CONCLUSIONS: Clinical characteristics and outcomes did not differ significantly between SCC and non-SCC. However, chemotherapy regimens differed to some extent, and the possibility of difference in overall survival among pathological types of non-SCC was suggested.

Adjuvant chemotherapy for a primitive neuroectodermal tumor of the uterine corpus: A case report and literature review.

A primitive neuroectodermal tumor (PENT) belongs to the category of a Ewing sarcoma. A PENT of the uterus is rare and has been known to be very aggressive by nature. Owing to the rarity of the tumor, there is no optimal treatment at present. In many cases, after hysterectomy, chemotherapy or radiation therapy has been performed. However, an effective chemotherapy regimen was unclear. In the soft tissue sarcoma area, the chemotherapy approach has recently greatly improved. Vincristine, doxorubicin, cyclophosphamide, ifosfamide and etoposide (VDC-IE) therapy has improved the survival rate of patients with Ewing sarcoma/PENT. Thus, VDC-IE therapy may be used for a uterine PENT. Here, we report a case of a uterine PENT in a premenopausal woman successfully treated with multimodality treatment including VDC-IE therapy and discuss the optimal chemotherapy for a uterine PENT through a literature review.

Thrombolysis with intravenous recombinant tissue plasminogen activator during early postpartum period: a review of the literature.

Thromboembolic events are one of the leading causes of maternal death during the postpartum period. Postpartum thrombolytic therapy with recombinant tissue plasminogen activator (rt-PA) is controversial because the treatment may lead to massive bleeding. Data centralization may be beneficial for analyzing the safety and effectiveness of systemic thrombolysis during the early postpartum period. We performed a computerized MEDLINE and EMBASE search. We collected data for 13 cases of systemic thrombolytic therapy during the early postpartum period, when limiting the early postpartum period to 48 hours after delivery. Blood transfusion was necessary in all cases except for one (12/13; 92%). In seven cases (7/13; 54%), a large amount of blood was required for transfusion. Subsequent laparotomy to control bleeding was required in five cases (5/13; 38%), including three cases of hysterectomy and two cases of hematoma removal, all of which involved cesarean delivery. In cases of transvaginal delivery, there was no report of laparotomy. The occurrence of severe bleeding was high in relation to cesarean section, compared with vaginal deliveries. Using rt-PA in relation to cesarean section might be worth avoiding. However, the paucity of data in the literature makes it difficult to assess the ultimate outcomes and safety of this treatment.