Cuff-less blood pressure estimation from photoplethysmography signal and electrocardiogram.

In recent studies, the physiological parameters derived from human vital signals are found as the status response of the heart and arteries. In this paper, we therefore firstly attempt to extract abundant vital features from photoplethysmography(PPG) signal, its multivariate derivative signals and Electrocardiogram(ECG) signal, which are verified its statistical significance in BP estimation through statistical analysis t-test. Afterwards, the optimal feature set are obtained by usnig mutual information coefficient analysis, which could investigate the potential associations with blood pressure. The optimized feature set are aid as an input to various machine learning strategies for BP estimation. The results indicates that AdaBoost based BP estimation model outperforms other regression methods. Concurrently, AdaBoost-based model is further analyzed by using the Histograms of Estimation Error and Bland-Altman Plot. The results also indicate the great BP estimation performance of the proposed BP estimation method, and it stays within the Advancement of Medical Instrumention(AAMI) standard. Regarding the British Hypertension Society (BHS), it achieves the grade A for DBP and grade B for MAP. Besides, the experimental result illustrated that our proposed BP estimation method could reduce the MAE and the STD, and improve the r for SBP, MAP and DBP estimation, respectively, which further demonstrates the feasibility of our proposed BP estimation method in this paper.

[Analysis of the diagnosis, treatment and prognosis in acute obstruction of proximal and distal colorectal cancers].

OBJECTIVE: The study aimed to review the treatment and prognosis of acute obstruction of colorectal cancers and to compare different treatment strategies of those cancers, and to evaluate the risk factors affecting perioperative complications. METHODS: Clinical data of 184 patients with acute obstruction of colorectal cancer undergone operation were analyzed retrospectively. RESULTS: A total of 184 patients with acute obstruction of colorectal cancer was collected in this study, including 58 patients with proximal and 126 patients of distal colorectal cancers. Perioperative death occurred in 2/58 patients (3.4%) with distal colorectal cancer and 6/126 cases (4.8%) of distal colorectal cancer (P > 0.05). The overall perioperative complications in the two groups were not significantly different (P = 0.794). Among the 58 patients with proximal colorectal cancer, one patient underwent colostomy, but among the 126 patients with distal colorectal cancer, 41 patients underwent colostomy, showing a significant difference between the two groups (P = 0.002). ASA scores (grade 3 - 4), elderly age (≥ 70 years) and colon perforation peritonitis were independent prognostic factors associated with perioperative mortality and morbidity. Patients in the self-expandable metallic stent (SEMS) group had a significantly shorter hospital stay (25.4 ± 8.3) d than that in the emergency surgery group (32.8 ± 16.4) d, (P = 0.039). CONCLUSIONS: Endoscopic stent implantation provides an acceptable modality of palliation for acute proximal large bowel obstruction caused by malignancies. In acute colorectal cancer obstruction, SEMS can provide a minimally invasive management compared with surgical intervention.

[Comparison of the diagnosis and treatment of mechanical bowel obstruction due to tumor or other causes].

OBJECTIVE: The aim of this study was to review the etiology and pathogenesis of patients who underwent surgery for mechanical bowel obstruction. The treatment and prognosis of bowel obstructions caused by intra-abdominal tumors were compared with those due to other causes. METHODS: The clinical data of 203 patients with mechanical bowel obstruction undergoing operation were analyzed retrospectively. The tumor cases were classified as group I, and all other cases as group II. A range of factors were investigated to estimate the postoperative outcome: gender, age, comorbidities, symptoms and findings of physical and radiological examinations, sites of the obstruction, etiology, therapeutic approach, postoperative complications and mortality. RESULTS: Group I included 73 patients and Group II 130. Large bowel carcinoma and peritoneal adhesions were the most common causes of Group I and II, contributing 58 and 86 of all cases, respectively. There was no significant difference in terms of gender between the two groups, but the rate of elderly (≥ 70 years) patients was significantly higher (53.4%) than that of the < 70 years old patients (35.4%) (P = 0.012). There was a significant difference between the patients with previous surgical operation history in the tumor group (23.3%) and non-tumor group (58.5%) (P < 0.001). In the 73 cases of the tumor group, the obstruction was located in the large bowel in 58 cases (79.5%), small bowel in 12 cases (16.4%), both small and large bowels in 2 cases (2.7%) and gastric cancer invading the splenic flexure of colon in 1 case, while in the non-tumor group, 111 cases (85.4%) of the obstruction was located in the small bowel and 19 cases (14.6%) and in the large bowel (P < 0.001). Sixty-six cases (90.4%) of the tumor-group underwent intestinal segment excision and 21 cases (28.8%) underwent intestinal fistulation in the tumor group, but in the non-tumor group 61 cases (46.9%) underwent intestinal segment excision and 5 cases (3.8%) underwent intestinal fistulation (all P < 0.001). The hospital stay was (18 ± 6) days in the tumor group and (11 ± 3) days in the non-tumor group (P < 0.01). The complication rate (P = 0.104) and mortality rate (P = 0.187) were not significantly different between the two groups. CONCLUSIONS: Tumor mechanical bowel obstruction is more frequently seen in patients in elder age, with colorectal location and without previous operation history. CT scan may provide effective diagnosis and ascertain the presence of the malignant obstruction. Intestinal fistulation is more often needed in patients with tumor intestinal obstruction and endoscopic stenting is a safe option in selected patients with tumor intestinal obstruction.

[Analysis of risk factors affecting operative outcome of small bowel obstruction].

OBJECTIVE: To evaluate the risk factors affecting the early postoperative outcomes in patients with small bowel obstruction. METHODS: Clinical data of 193 patients with small bowel obstruction undergone operation were analyzed retrospectively. A range of factors were investigated to estimate postoperative outcome, including gender, age, comorbidities, etiology of obstruction, presence of strangulated bowel (viable or nonviable), leukocyte count, temperature, and heart rate. Logistic regression analysis was used to study the prognostic value of each significant variable in terms of postoperation. RESULTS: The major causes of small bowel obstruction were adhesion and hernia, contributing 38.9% and 37.8% of all cases, respectively. Strangulation occurred in 42.0% and caused nonviable bowel in 23.3% of obstructing episodes. Elderly (>or=70 years), diabetes, malignant tumors WBC >15x10(9)/L were independent significant factors associated with bowel strangulation. The overall complication rate was 16.1%, the 30-day mortality was 4.1%, and the median postoperative hospital stay was 13 days. Age >or=70 years and bowel resection were significantly associated with postoperative complications in the univariate analysis. Only elderly and malignant obstruction were significantly associated with operative mortality in multivariate logistic regression. CONCLUSIONS: Surgery for small bowel obstruction is still associated with significant mortality and morbidity. Elderly is significantly associated with an increased incidence of strangulation, operative mortality, and complications.