1
|
Li S, Brandt M, Fensholt R, Kariryaa A, Igel C, Gieseke F, Nord-Larsen T, Oehmcke S, Carlsen AH, Junttila S, Tong X, d’Aspremont A, Ciais P. Deep learning enables image-based tree counting, crown segmentation, and height prediction at national scale. PNAS Nexus 2023; 2:pgad076. [PMID: 37065619 PMCID: PMC10096914 DOI: 10.1093/pnasnexus/pgad076] [Citation(s) in RCA: 3] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 08/24/2022] [Revised: 01/28/2023] [Accepted: 02/27/2023] [Indexed: 04/18/2023]
Abstract
Sustainable tree resource management is the key to mitigating climate warming, fostering a green economy, and protecting valuable habitats. Detailed knowledge about tree resources is a prerequisite for such management but is conventionally based on plot-scale data, which often neglects trees outside forests. Here, we present a deep learning-based framework that provides location, crown area, and height for individual overstory trees from aerial images at country scale. We apply the framework on data covering Denmark and show that large trees (stem diameter >10 cm) can be identified with a low bias (12.5%) and that trees outside forests contribute to 30% of the total tree cover, which is typically unrecognized in national inventories. The bias is high (46.6%) when our results are evaluated against all trees taller than 1.3 m, which involve undetectable small or understory trees. Furthermore, we demonstrate that only marginal effort is needed to transfer our framework to data from Finland, despite markedly dissimilar data sources. Our work lays the foundation for digitalized national databases, where large trees are spatially traceable and manageable.
Collapse
Affiliation(s)
- Sizhuo Li
- To whom correspondence should be addressed: ;
| | | | - Rasmus Fensholt
- Department of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen 1350, Denmark
| | - Ankit Kariryaa
- Department of Computer Science, University of Copenhagen, Copenhagen 2100, Denmark
| | - Christian Igel
- Department of Computer Science, University of Copenhagen, Copenhagen 2100, Denmark
| | - Fabian Gieseke
- Department of Computer Science, University of Copenhagen, Copenhagen 2100, Denmark
- Department of Information Systems, University of Münster, Münster 48149, Germany
| | - Thomas Nord-Larsen
- Department of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen 1350, Denmark
| | - Stefan Oehmcke
- Department of Computer Science, University of Copenhagen, Copenhagen 2100, Denmark
| | - Ask Holm Carlsen
- Department of Earth Observations, The Danish Agency for Data Supply and Infrastructure, Copenhagen 2400, Denmark
| | - Samuli Junttila
- Department of Forest Sciences, University of Eastern Finland, Joensuu 80101, Finland
| | - Xiaoye Tong
- Department of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen 1350, Denmark
| | | | | |
Collapse
|
2
|
Graversen DS, Christensen MB, Pedersen AF, Carlsen AH, Bro F, Christensen HC, Vestergaard CH, Huibers L. Safety, efficiency and health-related quality of telephone triage conducted by general practitioners, nurses, or physicians in out-of-hours primary care: a quasi-experimental study using the Assessment of Quality in Telephone Triage (AQTT) to assess audio-recorded telephone calls. BMC Fam Pract 2020; 21:84. [PMID: 32386511 PMCID: PMC7211335 DOI: 10.1186/s12875-020-01122-z] [Citation(s) in RCA: 7] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 07/01/2019] [Accepted: 03/04/2020] [Indexed: 11/10/2022]
Abstract
Background To explore and compare safety, efficiency, and health-related quality of telephone triage in out-of-hours primary care (OOH-PC) services performed by general practitioners (GPs), nurses using a computerised decision support system (CDSS), or physicians with different medical specialities. Methods Natural quasi-experimental cross-sectional study conducted in November and December 2016. We randomly selected 1294 audio-recorded telephone triage calls from two Danish OOH-PC services triaged by GPs (n = 423), nurses using CDSS (n = 430), or physicians with different medical specialities (n = 441). An assessment panel of 24 physicians used a validated assessment tool (Assessment of Quality in Telephone Triage - AQTT) to assess all telephone triage calls and measured health-related quality, safety, and efficiency of triage. Results The relative risk (RR) of poor quality was significantly lower for nurses compared to GPs in four out of ten items regarding identifying and uncovering of problems. For most items, the quality tended to be lowest for physicians with different medical specialities. Compared to calls triaged by GPs (reference), the risk of clinically relevant undertriage was significantly lower for nurses, while physicians with different medical specialties had a similar risk (GP: 7.3%, nurse: 3.7%, physician: 6.1%). The risk of clinically relevant overtriage was significantly higher for nurses (9.1%) and physicians with different medical specialities (8.2%) compared to GPs (4.3%). GPs had significantly shorter calls (mean: 2 min 57 s, SD: 105 s) than nurses (mean: 4 min 44 s, SD: 168 s). Conclusions Our explorative study indicated that nurses using CDSS performed better than GPs in telephone triage on a large number of health-related items, had a lower level of clinically relevant undertriage, but were perceived less efficient. Calls triaged by physicians with different medical specialities were perceived less safe and less efficient compared to GPs. Differences in the organisation of telephone triage may influence the distribution of workload in primary and secondary OOH services. Future research could compare the long-term outcomes following a telephone call to OOH-PC related to safety and efficiency.
Collapse
Affiliation(s)
- D S Graversen
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark. .,Department of Public Health, Aarhus University, Bartholins Allé 2, 8000, Aarhus C, Denmark.
| | - M B Christensen
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark.,Department of Public Health, Aarhus University, Bartholins Allé 2, 8000, Aarhus C, Denmark
| | - A F Pedersen
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark.,Department of Clinical Medicine, Aarhus University, Aarhus, Denmark
| | - A H Carlsen
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark
| | - F Bro
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark.,Department of Public Health, Aarhus University, Bartholins Allé 2, 8000, Aarhus C, Denmark
| | - H C Christensen
- Emergency Medical Services, Copenhagen, Denmark.,The National Clinical Databases (RKKP), Copenhagen, Denmark
| | - C H Vestergaard
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark
| | - L Huibers
- Research Unit for General Practice, Aarhus, Bartholins Allé 2, 8000, Aarhus C, Denmark
| |
Collapse
|
3
|
Simmons RK, Carlsen AH, Griffin SJ, Charles M, Christiansen JS, Borch-Johnsen K, Sandbaek A, Lauritzen T. Variation in prescribing of lipid-lowering medication in primary care is associated with incidence of cardiovascular disease and all-cause mortality in people with screen-detected diabetes: findings from the ADDITION-Denmark trial. Diabet Med 2014; 31:1577-85. [PMID: 25185778 PMCID: PMC4283975 DOI: 10.1111/dme.12574] [Citation(s) in RCA: 11] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/11/2014] [Revised: 06/16/2014] [Accepted: 08/25/2014] [Indexed: 12/03/2022]
Abstract
AIMS To examine variation between general practices in the prescription of lipid-lowering treatment to people with screen-detected Type 2 diabetes, and associations with practice and participant characteristics and risk of cardiovascular events and all-cause mortality. METHODS Observational cohort analysis of data from 1533 people with screen-detected Type 2 diabetes aged 40-69 years from the ADDITION-Denmark study. One hundred and seventy-four general practices were cluster randomized to receive: (1) routine diabetes care according to national guidelines (623 individuals), or (2) intensive multifactorial target-driven management (910 individuals). Multivariable logistic regression was used to quantify the association between the proportion of individuals in each practice who redeemed prescriptions for lipid-lowering medication in the two years following diabetes diagnosis and a composite cardiovascular disease (CVD) outcome, adjusting for age, sex, prevalent chronic disease, baseline CVD risk factors, smoking and lipid-lowering medication, and follow-up time. RESULTS The proportion of individuals treated with lipid-lowering medication varied widely between practices (0-100%). There were 118 CVD events over 9431 person-years of follow-up. For the whole trial cohort, the risk of CVD was significantly higher in practices in the lowest compared with the highest quartile for prescribing lipid-lowering medication [adjusted odds ratio (OR) 3.4, 95% confidence interval (CI) 1.6-7.3]. Similar trends were found for all-cause mortality. CONCLUSIONS More frequent prescription of lipid-lowering treatment was associated with a lower incidence of CVD and all-cause mortality. Improved understanding of factors underlying practice variation in prescribing may enable more frequent use of lipid-lowering treatment. The results highlight the benefits of intensive treatment of people with screen-detected diabetes (Clinical Trials Registry No; NCT 00237549).
Collapse
Affiliation(s)
- R K Simmons
- MRC Epidemiology Unit, University of Cambridge School of Clinical Medicine, Cambridge, UK
| | | | | | | | | | | | | | | |
Collapse
|
4
|
Maindal HT, Carlsen AH, Lauritzen T, Sandbaek A, Simmons RK. Effect of a participant-driven health education programme in primary care for people with hyperglycaemia detected by screening: 3-year results from the Ready to Act randomized controlled trial (nested within the ADDITION-Denmark study). Diabet Med 2014; 31:976-86. [PMID: 24646371 DOI: 10.1111/dme.12440] [Citation(s) in RCA: 9] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/15/2013] [Revised: 11/08/2013] [Accepted: 03/13/2014] [Indexed: 11/29/2022]
Abstract
AIM To assess whether a 12-week participant-driven health education programme offered to individuals with screening-detected hyperglycaemia in Danish primary care would lead to improvements in cardiovascular risk factors, health behaviour and patient-reported outcomes after 3 years. METHODS We conducted a randomized controlled trial in 509 patients with screening-detected hyperglycaemia (impaired fasting glucose, impaired glucose tolerance or type 2 diabetes) from 33 general practices in Denmark. Individuals were pre-randomized to receive (i) routine care (n = 187), or (ii) an invitation to participate in the Ready to Act health education programme (n = 322). The programme was delivered over 12 weeks in primary care and focused on motivation, action experience, informed decision-making and social involvement to promote health behaviour change. The primary outcome was 10-year modelled cardiovascular risk. RESULTS Of 322 individuals, 123 (38%) received the intervention and 436/509 individuals (86%) returned for follow-up assessment. There was no difference between the trial groups in modelled cardiovascular risk at 3 years (relative difference: 1.01; 95% CI: 0.84 to 1.23). Total cholesterol was lower (-0.24mmol/l, 95% CI: -0.45 to -0.03, P = 0.027), and patient activation was higher in the intervention than in the control group (5.3, 95% CI: 0.97 to 9.7). No other between-group differences were observed for any cardiovascular risk factor, health behaviour or patient-reported outcome variables. Subgroup analyses suggested that the intervention was more beneficial in those with impaired fasting glucose/impaired glucose tolerance than in those with type 2 diabetes. CONCLUSION For patients with screening-detected hyperglycaemia, a participant-driven health education programme was not associated with improvements in most clinical, behavioural and patient-reported outcomes after 3 years of follow-up.
Collapse
Affiliation(s)
- H T Maindal
- Health Promotion and Health Services Research, Department of Public Health, Aarhus University, Aarhus, Denmark
| | | | | | | | | |
Collapse
|